U.S. Army Signals Intelligence in World War II: A Documentary History (CMH pub)

u.s. ARMY SIGNALS INTELLIGENCE IN WORLD WAR II A DOCUMENTARY HISTORY

Edited by James L. Gilbert and John P. Finnegan

MLfU.RY frjSfRYClIOfl

CENTER OF MILITARY HISTORY UNITED STATES ARMY WASHINGTON, D.C. , 1993

Library of Congress Cataloging-in-Publication Data U.S. Anny Signals Intellige nce in World War II : a documentary hi story / edited by James L. Gilbert and John P. Finnegan. p. em. I. World War. 1939-1945- Military intelligence-United StatesHi story-So urces. 2. World War. 1939- 1945--Cryptogra ph y-Hi storySources. 3. United States. Signal Security Agency-HistorySources. I. Gilbert, James L. (James Leslie). 1943II. Finnegan, John Patrick. D8 10.s7U7 1993 940.54 '85 '0973--<1c20 92-43143 CIP

CM H Pub 70-43

First Printing For S
Superintendcnt of Documcnls. Mail Stop: SSGP. Washington. DC 20402-9328

ISBN 0-16-037816-8

Foreword This book is part of the Anny historical community's commemoration of the fiftieth anniversary of World War II. American victory in that conflict was brought about not only by the valor of our fightin g men and the immensity of our productive capacity, but also by the availability of superb military intelligence. Much of this intelligence came from the ability of our armed forces to intercept and decipher the most sec ret communications of their adversaries. For many years security considerations prevented any public mention of these successes in the official histories . Now much of the story can be told. To preserve the memory of the Army's role in this intelligence war, the U.S. Anny Center of Military History has joined with the History Office, U. S. Anny Intelligence and Security Command (INSCOM), to publish thi s collection of documents on Army signal s intelligence in World War II . INSCOM carries on the heritage of the Army 's World War [I Signal Security Agency, which by breaking the Japanese diplomatic ciphers and military codes helped speed the way of our forces to victory. The book is intended both for an Army audience and for the general public- including those World War II veterans who participated in the signals intelligence war and who for so many years were constrained to keep their contributions secret. The security barriers have now been lifted , and the Army is proud to acknowledge those contributions.

CHARLES F. SCANLON Major General , USA Commanding General U.S. Army Intelligence and Security Command

HAROLD W. NELSON Bri gadier General , USA Chief of Military Hi story

II I

Preface The Army 's contributions to the success of the American communication s inte lli gence-COMINT-effort in Wor ld War II are not widely known. This volume represents a modest attempt to close this infonnation gap. The core of the book consists of World War II-era documents generated by various cryptologic organizations of the U.S . Army. Originally, most of these documents were highly c lass ified. As a result of initiatives undertaken in the 1970s, they have now been re leased to the National Archives and Records Administration in the form of Special Research Hi stories after declassification by the National Security Agency. The documents included necessarily represent only a small fraction of the material that is now in the public domain. It is hoped that the publication of thi s book will he lp to alert the public to the ex istence of thi s relatively unexp lored mass of hi storical documentation. Although these materia ls are being in creas ingly used by profess ion al hi storians, only two co llections of these documents have yet been publi shed. In 1977, the Department of Defense released a multivolume co mpilation entitled The "Magic" Background of Pearl Harbor; more rece ntl y, Dr. Ronald Sp ector ha s edited Listening to th e Enemy: Key Documents on the Role o{ Communications Intelligence in the War with Japan . We can assure the reader that the present volume does not ex haust the possibilities of further ex ploiting this largely untapped source. The documents themselves consist for the most part of official hi stories and government memoranda. For the sake of variety, we have also included one highly unofficial hi story and examples of intelli gence reports and intercepted messages. Some of the documents have been rather heav il y san iti zed be fore release, as is indicated by the [text wi thheld] not ati ons libera ll y sprinkled through them. It should be pointed out that the declassification/ saniti zation process went on fo r a number of years (and is still continuing), and that guidelines as to the releasability of material have changed during thi s period. As a result, the careful reader will note certain inconsistenc ies. The ed itors of thi s volume have made no attempt to reverse any of the original declassification decisions. They have, however, not hesitated to abridge documents that were overly long or treated subjects repetitive ly. The editors have also chosen to eliminate those portion s of documents that had been so

v

heavily sanitized as to render them less than intelligible. Those who are curious about what we may have omitted may check the original documents at the National Archives and Records Administration. It should be noted that most typographic mistakes in the originals have been retained. No book comes into existence without some type of collaborative process going on , and this one is no exception. We would like to begin by thanking Col. James H. P. Kelsey, Deputy Chief of Staff, Operations, U.S. Army Intelligence and Security Command, for the unfailing support he has provided to the History Office in all of its projects. The manuscript itself was prepared and proofread by Ms . Diane L. Hamm, the History Office writer-editor. John Elsberg of the U.S. Army Center of Military History did much to enhance the final product through astute editing and wise counsel. John Greenwood, Chief of the Center of Military History's Field and International Division, made arrangements for an earlier version of the introductory essay to be presented at the 1990 Army Historians Conference. John Birmingham designed and formatted the final product, working closely with the authors and the editoria l staff. Finally, we would be remiss if we did not thank the numerous declassification specia li sts at the National Security Agency whose dedicated work made the publication of this book ultimately possible. Fort Belvoir, Virginia

JAMES L. GILBERT JOHN P. FINNEGAN

VI

Contents Chapter

Page

I. U.S. ARMY SIGNALS INTELLIGENCE IN THE

SECOND WORLD WAR: AN OVERVIEW.........................

3

II. THE ROAD TO PEARL HARBOR ........... ...... ...... ... ........ ........... ... ...... SRH 29: A Brief History of the Signallntelligence Service (excerpt, 1945)................................ ........................................... I. Riverbank Laboratories............................................... 2. Cryptanalysis in World War 1..... ....... .... ... ... ..... ... ... .... . 3. Relocation of MI-8...................................... ........ ........ 4. Radio Intercepl............................. ............................... 5. Yardley's Organization................ ........... ... ................ .. 6. Deficiencies of Interwar Organization ... .... .. ..... ........ .. 7. Impact of Secretary Stimson....................................... 8. The End of Yardley .......... ..... ... ........ ... ... ................. ... . 9. The Beginnings of the SIS .......................................... 10. Navy Efforts... ............................................................. II. Conclusion .................................................................. SRH 134: Expansion of the Signal Intelligence Service from 1930 t07 December 194 1 ( 1945) ........................... ......... SRH 135: Hi story of the Second Signal Service Battalion, Appendix; Memorandum to Adjutant General: Signal Intelligence ( 1938) ........................................................ "Magic" Background of Pearl Harbor. Volume IV. Appendix. Washington-Tokyo Message #812.. ...... ... ........ ... ........... ...... ..... SRH 11 5: U.S. Army Investigations Into the Handling of Certain Communications Prior to the Attack on Pearl Harbor: Testimony of Brig. Gen. Hayes W. Kroner (excerpt, 1944) .....

17 18 18 18 19 20 21 22 23 24 25 25 26 26

37 39

40

Note: SRH numbers refer to Special Research Histories: declassified/sanitized cryptologic documents released to the National Archives and Records Administration by the National Security Agency.

vii

Chapter

Page

III. ACCEPTING THE CHALLENGE .................... ............... ..... ...........

47

SRH 116: Origins, Functions and Problems of the Special Branch, M.I.S. ( 1943)................. ...................................... ....... 48 Responsibilities of War Department for Intelligence Derived from Cryptanalysis......................................... 48 Handling of Intercept Intelligence Prior to War................. 49 Decision of the Secretary of War Concerning Exploitation of Cryptanalytic Ficld ............................. 49 Organization of MIS Section to Handle Intercepts............ 50 Organization of Special Branch ..................................... .... 50 Work of the Special Branch ............................................... 51 Increase in Volume of SSD Material........ ... ........ ... ... ......... 54 Basic Factors in the Problem of Cryptanalytic Intelligence .................. ...... ..................................... ..... 55 Personnel Situation of the Special Branch Officers and Civilians .................................................. 60 SRH 141 -2: Papers from the Persona l Files of Alfred McComlack, Colonel. AUS, Spec ial Branch , G-2, Military Intelligence Division , War Department General Staff; Memorandum for Colonel Clarke: Personnel Problems, B Section. Special Branch (1944) ........................... 63 IV. TRAINING THE FORCE .................................................................

67

SRH 358: History of the Training Divi sion, Signal Security Agency; Hi story of Schooling of Commissioned Officers in the Vint Hill Farnls School ( 1945) .......................... ...... ....... 10 March 1942 to 5 October 1942..................................... 5 October 1942 to Dale ................................. ............ ......... "Third Battle of Manassas" (excerpt, 1944) ................................... They Fell Out in Front of the Orderly Room........ .. ........... Admini stration for All Concerned .............. ...... ......... ........ Education in Darkest Virginia.............. ... ............. ..............

68 68 69 72 72 77 80

V. CRACKING THE CODES............................... .. ...............................

87

Achievements of the Signal Security Agency in War II (excerpt, 1945)............. ...................................... Building the Organization .................... ......... ................ The Production of [nfonnation.......................... .

88 88 94

SRH 349: World I. II.

V III

Chapter

Page A. Interception....................... ........... .............. ............. B. Traffic Analysis ..................................................... C. Cryptanalysis .......................................................... D. Solutions .................................................... ...... ... .... I. Japanese Diplomatic and Military Attache Traffic ................................................ 2. Japanese Amly and Air Force Traffic ............... 4. German Army and Air Force Traffic ................ 8. Commercial Code Traffic ................................. 9. Special Problems .............................................. E. Some General Remarks ........................................... Central Bureau History (1950) ........................................................ SRH 280: An Exhibit of the Important Types of Intelligence Recovered Through Reading Japanese Cryptograms (excerpt). Berlin-Tokyo Message #756, 1944 .......................... I. Weapons Connected with the Army .............................. 2. Weapons Connected with the Navy .............................. 3. Weapons Connected with the Air Force........................ 4. Equipment for Long Distance Bombardment ...............

94 96 97 99 100 101 108 110 110 110 112

113 114 115 I 15 116

VI. PRODUCING INTELLIGENCE..................................................... 117 SRH 185: War Experience of Alfred McCormack (excerpt, 1947).... ...... ...... ...... ...... .. .. .. .............. ...... ..................... I 18 SRH 141 - 2: Papers from the Personal Files of Alfred McCormack; Memorandum for Mr. McCloy: Personnel Situation of Special Branch, M.I.D. (excerpt, 1943) ................ 128 SRH 005: Use of MX/CSS Ultra by War Department; European Summary of II April 1945 ....................................... 130 VII. DISSEMINATING THE PRODUCT ....................................... ... .. . 135 SRH 026: Marshall Letter to Ei senhower (1944) ........................... 136 SRH 033: History of the Operations of Special Security Officers Attached to Field Commands, 1943- 1945................. 139 SRH 031: Trip Reports- ULTRA in the Mediterranean; Report of Lt. Col. Rood (1944) ........................ ........................ 145 SRH 022 : ULTRA and the U.S. Seventh Army .............................. 150 Introduction ..................................... .............. ........................... 150 Operations.......................... .. ........................................... ... ....... 150 Security ........... ......................................................................... . 152

IX

Chapter

Page

Conclusions ................................... ... ... .... ........... ......... 152 Tab A .. .............. ........................................................... ....... 153 SRH 032 : ULTRA in the South west Pacific Area (SW PA) ............ 155 VTII . T HE POLIT ICS OF COMINT .............................................. ..... .... 159 SRH 14 1- 1: Papers fro m the Personal Fi les of Alfred McCormack; Memorandum for Mr. McCloy : Signal Intelligence Service ( 1942) ........... .......................... .................. SRH 14 1- 2: Papers from the Personal Files of Alfred McCormack; Memorandum for General McNamey ( 1944) .......................... .. ................................. .. ....... .. ... ...... SRH 14 1- 2: Papers from the Personal Files of Alfred McCormac k; Memorandu m for General Bissell : Arm y-Navy Ag reement Regarding ULTR A ( 1944) .. ............... A. Backg round of Arm y- Navy Relations in Ultra Fi eld .... B. Proposed Restrictions .. .................................................. C. Proposed Scope of Materi al to Be Released by Navy ... D. Proper Interpretati on of Agreement .. .. .......................... E. European Intell igence .. .................... .. .... .. ...................... SRH 043: Comments on Marshall-Dewey Exchange ( 1944) ........ IX. TACT ICA L COM INT.. ........ ...... .................................................... SRH 124: O perational Hi story of the 849th Signal Intelligence Serv ice (excerpt , 1945) .......... .. .. .... .. .............. .. ......................... I. Beginnings .... .. ...................... .. .................................. .. ... General ............................ ...... ...... ....... ............... .......... II . Intell igence Branch .. .. .. ...... .. ........... .. .... .. ...................... Genera l Functions.............. ...... ... .. ... ......... .. ................. I. Introducti on .................................... .. .......... .. ..... 2. Cryptanalysis ........................................ .. ... .. ..... 4. Laboratory Fac il ities .... .. .......... .. .. .. .... .. .. ........... Major Problems Encountered .......... .. .. .... .. .. ................ 2. U.S. Signal Radio Inte lligence Companies ....... 3. Proc urement and T raining of Personne l ........... 4. Organi zation and Eq uipment of Field Detachments.. ............ .................... .. .. .. .. ...... .. . SRH 048: Summary of Operational Activity, S ignal Security Detac hment " 0 ", Covering the Period I September 1944 to I Apri l 1945 .. .... .. .................... .. ................. .... .. .... .. .. .. ..

x

160

163

166 166 168 168 169 170 17 1 179 180 180 180 184 184 184 185 187 187 187 189 190

192

Chaprer

Page

I. Purpose .. ....... ... ... ........... .............................................. II. Miss ion of the Detac hment ......................... .............. .. Ill . Signal Intell igence Di vision, Headqua rters, ETOUSA ..................................................................... IV G-2 Section, Tactical Headquarters, 12th Anny Group .. .. .... .............. .. ...... .. .... .................... V Signal Rad io Inte lligence and Signal Security Companies, 12th AmlY Group .. .. .. ........................ VI. Flanking Army Groups .......................................... .. ... SRH 042: T hird Army Rad io Intell igence History (excerpt, 1945) .................................................... .. .... .. .. .. ...... .. .. Operational Organi zati on and Proced ure of Radio Intelligence Agencies .............................................................. A. Organization .......... .. ......... ............................................. B. Radio Intelligence Procedu res ................ .. ..................... I. Operati onal ................................ .... .................. .. .. .. .. 2. Communi cat ions Set-Up .. .. .......... .. .............. .. ...... .. . C. Captu red Documents .... .. .. ............................................ D. Prisoner of War Interrogations ...................................... SRH 227: History of the 126th Signal Rad io Intelligence Company (excerpt, 1945) .. ................................................ ........ SRH 049: Technica l Signal Intelligence Transmitted Directly to G-2, 12th Aml y Group, ETO From 4 August 1944 to May 1945 (excerpt , 1945) ........................................................ X. RES HAPING T HE TOOLS .................... .. ............ .. .............................. SR H 169: Centrali zed Cont rol of U.S. Ann y Signal Inte lli gence; Memorandum for Chief of Staff: Centralized Cont ro l of Signal Intelligence Acti vities for War Against the Japanese ( 1945) ................................ .. ........... I. C ircumstances Leading to Proposal ............................. II . Proposa l to T heaters and Rep lies Thereto .................... lIT. Basic Reasons for Proposal. ......................................... Ava ilability of Traffic for Intercepti on ................. Cryptanalys is ........................................................ Usefulness of Results Obtained .. .. .. .... .. ...... .......... " Low Level" Signals Intelligence ............ .. ........... IV O bjections Raised by SWPA .. .. ................................... V. Procedure for Pl acing Centrali zed Control into Effect ...................................................... .. .. .... .. ...

XI

192 192 192 193 194 195 196 196 196 197 198 203 204 204 205

2 11 2 13

2 14 2 15 2 16 2 16 2 16 2 17 2 17 2 18 2 18 220

APPEND IXES: A. Chronology .............. ... .... .. ......... .... .... .. .... .. ....................................... 225 B. G lossary ........................................................... ... ... ... ...... ..... ...... ... .... 227 C. Dictionary of People, Places. and Terms .......................................... 233

ILLUSTRATIONS The following ilillstrations appear between pa[;es 112 and 113: Herbert O. Yardley Yardley 's New York Headquarters William F. Friedman Friedman and Hi s SIS Staff The Munitions Building 2d Signal Service Company Member on Hawaii The Ballieship Ari:ona Sinking at Pearl Harbor Henry L. Stimson John J. McCloy George Y. Strong Graduation Parade at Vint Hill Farms School Japanese Language Training Course Arlington Hall. Virginia Anny Signal Security Agency Staff Heads Army Cryptanalysts at Arlington Hall Calculating Machines at Arlington Hall

The following illllstrations appear beIWeen pages 166 and 167: Brig. Gen. Carter W. Clarke with Col. Alfred W. McCormack German Signal Troops The Briti sh Government Code and Cypher School Brig. Gen. Telford Taylor with Hi s Staff at Nuremberg General Dwight D. Eisenhower Addresses Paratroopers Genera l Douglas MacArthur German U-boat General George C. Marshall with Governor Thomas E. Dewey 21 st Signal Service Company's Radio Imercept Platoon Signal Security Derachmenl " D" Analysts 128th Signal Radio Inte lligence Company Imercept Personnel Radio Direction-Finding Site at Bri sbane, Au stralia Genera l Eisenhower, Maj. Gen. Hoyt S. Vandenben , Brig. Gen. Preston W. Cordennan, Col. Frank B. Rowlell , and Me Friedman

XII

u.s. ARMY SIGNALS INTELLIGENCE IN WORLD WAR II

CHAPTER I

u.s. Army Signals Intelligence in World War II: An Overview by John P. Finnegan The United States Anny entered World War II with an effective, functioning organization for the production of communications intelligenceCOM INT. Thi s was the Signal Intelligence Service (S IS ) of the Army Signal Corps. The SIS had been formed in 1930 to consolidate all Army cryptologic functions under the Signal Corps. It was a unified organization responsible for producing the Army 's own codes and cipher devices and for attempting to decrypt the communications of America's potential adversaries. Previously, the cryptanalytic function had been carried out by the War Department's Military Intelligence Division through a clandestine element based in New York, headed by Herbert O. Yardley, and funded with State Department support. However, thi s had been discontinued in 1929, partially as a result of Secretary of State Henry , L. Stimson 's conviction that "Gentlemen do not read each other's mail." The SIS was an organization shaped by its first chief, the legendary William F. Friedman. Friedman, who had previously served as the Signal Corps ' code and c ipher consultant , quickly recruited a small but extremely tal ented civilian staff. Even after a military officer was placed in charge of the SIS in 1935, Friedman continued to exercise a wide influence. During the period before World War II , the SIS published a se ries of influential studies on cryptology and pioneered the development of machine ciphers of unparall e led sop histication and security to safeg uard Army communications. It also implem ented trainin g courses to ed ucate reserve officers in cryptology, thus selling up a limited mobilization base for wartime expan-

4

SIGNALS INTELLIGENCE

sion . Another accomplishment was the establishment of a network of radio intercept stations, something which the Army had never attempted in peacetime; Yardley's earlier cryptanalytic efforts had depended on the cooperation of friendly cable companies to provide the raw material. ' The greatest prewar accomplishment of SIS, however, was its success in deciphering the machine cipher used by the Japanese Foreign Office for its most secret messages. Through purely cryptanalytic methods, the SIS was able to build an analog of this machine and read Japanese dispatches as quickly as the intended recipients. As a result, the ArnlY became privy to the thought processes that guided Japanese diplomacy. The secret breakthrough was shared with SIS 's Navy counterpart, OP-20-G , and later with the British. Otherwise, it was closely held. Decrypts of enciphered Japanese messages, code-named MAGIC, were passed on to a tightly restricted group of individuals in the War and Navy Departments and further disseminated to top policymakers (the Secretaries of War, Navy, and State, the President 's military adviser, and the President himself) . 3 Unfortunately, MAGIC could not ward off Pearl Harbor. The high-level communications of the Japanese Foreign Office did not carry the details of Japanese military planning. Consular traffic that did contain military intelligence indicators traveled in less secure systems, but it could not be translated in a timely fashion because of other priorities. Moreover, the very secrecy with which MAGIC was handled impaired its utility. A few upper-level officials received tantalizing snippets of information from critical messages, but the material itself was never subject to detailed exploitation and analysis over the long haul. As Friedman would later put it, "each message represented only a single frame, so to speak, in a long motion picture film.'" But no one had ever been in a position to watch the whole movie. The shock of the Pearl Harbor attack immediately led to the adoption of a more systematic approach to handling diplomatic intercepts. A first step was taken when the Secretary of War, Henry L. Stimson, directed that a Special Section be formed within the Military Intelligence Division 's Far Eastern Section. (The head of the Far Eastern Section had previously func tioned as the sole conduit of MAGIC.) Following an outside study conducted at Stimson's direction by a Chicago attorney, Alfred McCormack , the section was expanded to the status of a Special Branch within the Military Intelligence Service that was created as the operating arm of the Military Intelligence Division in March 1942. The Special Branch was headed by Col. (later Brig. Gen.) Carter Clarke; McCormack himself accepted a

OVERVIEW

5

colonel's commission and became Clarke's deputy. This organ izationa l innovation for the first time allowed sufficient numbers of ana lysts access to COMINT in ways that would allow them to use it in its entirety to build up an intelligence picture. ' Meanwhile, the Signal Intelligence Service was conducting a rapid expansion. In March 1942 the Military Intelligence Division (MID) recommended to the Chief Signal Officer that SIS move from its existing quarters in the Munitions Building in downtown Washington to new quarters that would provide greater security. MID also suggested that primary monitoring stations be estab li shed , one on each coas!. At the time, the seven smal l existing stations manned by the 2d Signal Service Company- the SIS collection unit- were not well positioned for wartime needs; the station nearest to the scene of action, Fort McKinley in the Philippines, had already been overrun by the Japanese Army. 6 Under wartime conditions, these steps were taken with a lacrity. In June 1942 the first Army elements took possession of Arlington Hall, a former women's junior college just outside Washington. That same month, personnel from the 2d Signal Service Battalion (as the intercept arm of the SIS was now designated) began operations at Vint Hill Farms in Warrenton, Virginia, which had been selected as the site for Monitoring Station Number I. A second major field station was soon set up at Two Rock Ranch near Petaluma, Ca liforn ia. In the meantime, SIS itself was redesignated. The term "Signal Intelligence" was thought to be too provocative; SIS first became the Signal Security Division , then the Signal Security Service; and achieved its final wartime designation as the Signal Security Agency (SSA) in the summer of 1943. 7 The Signal Corps' cryptologic organization recruited a largely civilian work force and began a process of rapid expans ion. The original buildings at Arlington Hall were quickly outgrown. In the winter of 1942 and the spring of 1943, two large temporary bui ldings were hastily constructed at Arlington Hall to handle the overflow. Meanwhile, the agency moved in new directions even as it grew. In the spring of 1942 liaison with the British introduced it to the new and promising art of traffic analysis , through which intelligence could be derived from the externals of a message even if the message itself could not be decrypted. In June 1942 the agency was given exclusive responsibility for exploiting foreign diplomatic communications, a mission previously s hared with the Navy. It also dispatched a small detachment to Australia and sent one of Friedman 's top cryptologists, Col.

6

SIGNALS INTELLIGENCE

Abraham Sinkov, to command it. This would become the American nucleus of a completely independent COMINT effort conducted in support of General Douglas MacArthur's Southwest Pacific Area by Central Bureau Brisbane, a combined U.S.-Australian operation headed by Brig. Gen. Spencer Akin, MacArthur's signal officer and a former chief of SIS. 8 However, there were still large deficiencies in the Army's COMINT effort. The effort remained fragmented with intell igence production and collection in separate compartments. The Specia'l Branch of the Military Intelligence Service, responsible for COMINT evaluation, had no association with the cryptanalysts at Arlington Hall. The latter group worked for the Signal Corps , which itself was only one component of the Army Service Forces. As a result, McCormack commented, a situation existed in which the Military Intelligence Service was "simply taking in what the Signal Corps catches and tums in , leaving with the Signal Corps the responsibility for determining how much it will catch and tum in, and from what sources.'~ And the Signal Security Agency itself did not monopolize COMINT collection, since tactical signals intelligence units in the field worked for the supported commanders, not SSA. 10 Moreover, the Army's COMINT effort was hampered by a severe shortage of personnel throughout the whole system. This created bottlenecks within the Special Branch and the Signal Security Agency. Part of the problem resulted from military regulations. The Secretary of War had decreed that no officer under 28 years old could be in Washington after 31 January 1942. Additionally, the Army was reluctant to offer direct commissions to qualified civilians. The Civil Service Commission was equally obstructive. McCormack noted rather sourly that it seemed "unable to distinguish an intelligence organization from an office that routes and stamps papers, and therefore forces on us an organization chart that calls for quotas of morons and semimorons .... " The Signal Security Agency, with much larger personnel requirements, was even more vulnerable to such practices, and it was hampered by the low priority given to it as a component of Anny Service Forces. Paradoxically, the importance of its work , which might have been used to justify a higher personnel priority, was too secret to divulge. " Even more frustrating was the slow progress made by SSA in actually breaking any military codes. As previously noted , Japanese cryptosystems worked on principles entirely different from those used in Japanese diplomatic communications. The Japanese Army used bulky reenciphered codes, not a cipher machine. Cryptanalytic continuity was lacking . The Navy had

OVERVIEW

7

been able to monitor traffic produced by Japanese naval exercises in the 1930s, but the Army had never encountered Japanese military nets until Bataan. As a result, SSA analysts at Arlington Hall were not able to make an entry into a single Japanese Army system until April 1943, when SSA and MacArthur's Central Bureau Brisbane made an almost simultaneous breakthrough. The first usable intelligence was not derived from exploitation of this Japanese Water Transport code until June 1943. 12 The Army was equally frustrated in its attempt to work on German mili tary systems. In April 1942, SSA had suggested to the British that it be allowed to work on the Enigma cipher machine used in high-level communications of the Gem1an am1ed forces. It was generally understood that the Government Code and Cypher School (the main British cryptanalytic center at Bletchley Park) had achieved some success in this field. However, the British rebuffed the Army 's efforts either to secure Enigma decrypts or to work on the Enigma problem. This studied lack of cooperation-justified by the British on grounds of security and economy of effort- became even more grating when the Army became aware that the British were willing to share high-level COMINT with the U.S. Navy to meet the operational needs of fighting the U-boat war. 13 Finally, America 's first exposure to actual land combat in the Tunisia campaign that followed Operation TORCH revealed major deficiencies in the structure of the Anny's tactical signals intelligence organization. When the Army entered World War II, it was envisaged that only two types of tactical COMINT units would be needed: radio intelligence platoons organic to the divisional signal companies and signal radio intelligence companies assigned on a basis of one per field am1y. Since it was assumed that analysis and translation would be accomplished by small radio intelligence staffs at theater or field am1Y level, as had been the case in World War I, no analytic personnel were included in these units. " The test of combat quickly demonstrated the unsuitability of this structure. The divisional platoons, in exposed positions near the front, were in too great a danger of being overrun to practice effectively their specialty; the signal radio intelligence companies, as organized, were too large and unwieldy to be used to support a corps, which had no dedicated communications intelligence support but which seemed to be the level of command most in need of it. Central processing proved to be impracticable, and the staffs at army level were too limited to handle the volume of exploitable COMINT. At the same time, the communications practices of the European

8

SIGNALS INTELLIGENCE

members of the Axis (but not the Japanese) presented large volumes of material sent in clear or in low-level codes susceptible to timely forward exploitation. The Army was initially forced to turn to the British to bolster its units with analytic personnel. Later, a numbered theater intelligence service , the 849th Signal Intelligence Service, was formed to support American COMINT units in the North African Theater of Operations. " [n short, the Anny's successes in the COM [NT field up to the midpoint of the war were limited. Its continuing triumphs in the field of diplomatic communications had not been matched by corresponding success against the military systems in use by its major adversaries. The Army 's COMINT effort had been hampered by its own organizational deficiencies and lack of priorities . The Army had been rebuffed by the British in its attempt to obtain access to German COMINT; it had been largely baffled by the Japanese military systems. Commenting on the situation in early January 1944, McCormack concluded morosely that "except for whatever the Navy has been able to do, we have achieved no corresponding victory over the Japanese. To date, in the field of radio intelligence, we have met largely with failure. " " McConnack 's dissatisfaction at the beginning of 1944 was understandable. A resistive personnel system had created bottlenecks within both Arlington Hall and the Special Branch. Arlington Hall 's initial breakthrough against Japanese codes in the summer of [943 had created "a critical need for at least 2,000 [additional] persons ... " , but it had taken six month s to begin " the lengthy and difficult job of recruiting such personnel. .. ." In the meantime , 200,000 unread Japanese messages had piled up . 17 But McCormack drew too dark a picture. Already the situation was becoming brighter. Mid-1943 had been a watershed in the Army 's final successful exp loitation of communications intelligence. Two major developments took place at this time . The first was Arlington Hall 's success in breaking and exploiting the Japanese Water Transport codes. Thi s proved to be a springboard for the exploitation of other Japanese military systems. The Signal Security Agency, now steadily expanding, was able to throw masses of personnel and batteries of IBM machines against the Japanese military cryptosystems. The capture of codes by troops in the field enhanced the ongoing efforts of cryptanalysts. Soon after McCormack's pessimistic evaluation, Arlington Hall had thus mastered the main Japanese Army code, a breakthrough that produced a wealth of intell igence. '" The second development was of equal importance. In mid-1943 the

OVERVIEW

9

Bri ti sh finally decided to take the United States Army into their confidence and reveal the full story of their success against the Enigma and other highgrade German systems. Thi s was accompli shed by a liai son visit to London by a joint Special Branch-SSA team: McCormack, Fri ed man , and Lt. Co l. Telford Taylor of the Special Branch. All barri ers to free exchange suddenly fell. American cryptanal ysts were allowed to work at Bl etchley Park, a Special Branch liai son officer in London supervised the di spatch of perti nent traffic to MID in Washington , and Arlington Hall was given a share of research on the Enigma problem. In turn , the United States agreed to adopt and implement stringent British sec urity practices on the transmission of COMINT to the field . " The new availability of CO MINT from both American and Britis h sources created for the first time the problem of di sseminati on to the field. The problem was solved by the creation of a system of Special Sec urity Officers (SSOs) along the pattern a lready establi shed by the Briti sh. The SSOs would serve as re pre se ntativ es of Special Branch to th e maj or American commands in the field, functioning as transmitters, guardian s, and interpreters of ULTRA, as hi gh-l eve l COMINT was now called. The first three SSOs were sent to the Pac ific in December 1943. Shortly thereafter, the British agreed to allow American SSOs to support the American commanders in the European and North African- Mediterranean- Middl e East Theaters. (Pre viously, the British Spec ial Liai son Unit system had handled all such dissemination of COMINT.) A subsequent decision to release ULTRA down to the field army and numbered air force level resulted in an even greater expansion of the Special Security Officer system. By August 1944 the SSO system had been allotted 172 officers and 65 enli sted communi cations personnel. 2
10

SIGNALS INTELLIGENCE

Branch entity continued to supervise dissemination of COMLNT to the field.) " A few months later, the artificial barriers that had separated collectors of COMINT from evaluators also fell. The fact that the Signal Security Agency was just a small part of the Signal Corps had disturbed War Department intelligence officers ever since 1942. By 1944 the SSA had become the Military Intelligence Division 's most important single source, furnishing it with 70 percent of its diplomatic intelligence and 80 percent of its infonnation on the Japanese military. This made the exist ing situation even more unsati sfactory. In August 1944 Assistant Secretary of War John J. McCloy wrote pointedly to Lt. Gen. Joseph McNarney, the Army Deputy Chief of Staff, that "I do not believe it is good practice to have the MIS by general directive and the establishment of general policies to attempt to create a workable intelligence organization in the low echelons of Army Service Forces as they are now attempting to do .,,22 The Signal Corps clung tenaciously to SSA , but in the end it was forced to back down. On 15 December 1944, the Military Intelligence Division was given operational control over the Signal Security Agency, leaving the Chief Signal Officer responsible only for the agency's administration. " Meanwhile, SSA continued to expand. By the summer of 1944 5,100 civi lians and 2,000 military personnel were on duty at Arlington Hall. The intercept facilities of the 2d Signal Service Battalion, SSA's collection ann , were also built up. In the spring of 1944 new fixed stations were established at New Delhi, India; Asmara, Eritrea; Fairbanks and Amchitka, Alaska; and Fort Shafter, Hawaii . In the fall the steady advance of American forces allowed another field station to be established on the island of Guam . The battalion al so took over " listening posts" at Be llmore, New York, and Resada, California, that had been previou s ly operated by the Office of Strategic Services (OSS). The OSS had used the facilities to track Axis propaganda broadcasts; SSA converted them to security monitoring stations. 24 At the tactical level Army signals intelligence was strengthened by the creation of new types of units. As previousl y noted, the Anny's experience in conducting the COMINT war in North Africa and later in the Mediterranean had demonstrated deficiencies in the existing organization of field assets. Divisional signal radio platoons had proven unsatisfactory; the existing signal radio intelligence companies were cumbersome and could function effectively only when supplemented by analytic and lingui st perso nnel from a theater-leve l s ignal intelligence se rvice. The War Department's initial reaction had been to move radio intelligence platoons

OVERVIEW

II

up to the corps signal battalion. However, this measure only jammed a small number of intelligence personnel into a larger unit performing an unrelated function, resulting in obvious security and operational disadvantages. In the northern European theater, a happier solution was found. Small signal service companies , numbered in a sequence beginning with 3250, were formed from existing resources in theater to provide COMINT support to each corps. These units contained their own analytical personnel. The large signal radio intelligence companies continued to operate at army, army group, and theater level. " In the Far East, because of the nature of the target and the tactical circumstances, the structure of the Army's tactical signals intelligence organization followed a different pattern. The Japanese target array offered little opportunity for local exploitation, and large Army formations were not fielded until the Philippine campaign of late 1944. Consequently, the four signal radio intelligence companies in the Southwest Pacific Area operated as collection arms for MacArthur's Central Bureau Brisbane rather than supporting the army commanders. In the Pacific Ocean Area two signal radio intelligence companies worked in similar fashion on high-level codes for Arlington Hall. " In addition , the Army Air Forces (AAF) created its own specialized intercept units to carry out a mission previously performed by signals intelligence units of the ground Army. In 1944 the AAF organized radio squadrons, mobile, to support each numbered air force . These large, battalion-size units, which included AAF intelligence analysts, were capable of operating in separate detachments and conducting intercept against both continuous-wave (Morse) and voice transmissions. 2J By 1945, then, the Army had largely perfected its structure for collecting, evaluating, and disseminating communications intelligence. In this field, however, the Army never acted as an autonomous player. MIS and SSA were partners in a vast combined eFfort in which responsibilities were largely divided along theater and target lines, with the British focusing on the German problem, and the Americans on the Japanese. However, the U.S. Army was deeply involved in all theaters: American SSOs furnished intelligence derived by Bletchley Park to the American commanders they supported in Europe, and Army personnel were at three of the four great Allied production centers. The Army controlled Arlington Hall ; Central Bureau Brisbane continued as a combined U.S .-Australian effort; and the Army had a foothold at Bletchley Park. Only the Wireless Experimental

12

SIGNALS INTELLIGENCE

Center at Bombay, India, remained a purely British operation. 28 Despite these steady improvements, the Army's COMINT arrangements were still less than satisfactory in three areas . SSA was still under the administrative control of the Chief Signal Officer, and this took its toll on any claim for resources. The fact that Col. Preston W. Corderman, who commanded 10,000 men and women in his dual capacity as chief of SSA and commander of the 2d Signal Service Battalion, did not receive a brigadier general 's star until June 1945 reveals something of the dimensions of the problem. " Second, the Army's strategic and tactical signals intelligence assets were still not coordinated properly. Tactical COMINT units were trained by Army Ground Forces , not SSA, and they functioned under the direction of theater and numbered air force commanders, not SSA, despite the fact that this division of labor did not reflect the reality that the communications intelligence process was a seamless web. As one contemporary evaluation put it , "radio intelligence operations, their availability for interception , the cryptographic systems used , and the usefulness of the results derived therefrom , all have no relation to the considerations which are determinative in assigning command responsibility for combat operations against enemy forces. " .'0

Finally, cooperation between AmlY and Navy COMINT operations had been slow to arrive, despite the fact that OP- 20-G, the Navy equivalent of SSA, was headquartered in another requisitioned junior college just a few miles away from Arlington Hall. An Army - Navy Communications Intelligence Coordinating Committee CANCICC) had been set up in early 1944, but had not been able to effect coordination all that well. The Navy had different uses for COMINT than the Army, and its security practices were more stringent. Navy hesitancy at sharing naval COMINT had been so great that McComlack at one time had toyed with the idea of tenninating all cooperation between the two services. In February 1945, however, the two service chiefs CArmy Chief of Staff General George C. Marshall and Navy Commander in Chief and Chief of Naval Operations Admiral Ernest W. Kin g) had agreed to set up a high - level Army - Navy Communication s Intelligence Board to finally coordinate a joint effort. " The new head of MID, Maj. Gen. Clayton Bissell , took this Navy initiative as an opportunity to resolve what he perceived as the remaining difficulties of Army signals intelligence organization. He protested that " the Army cannot participate on an inter-service project of this sort as long as its

OVERVIEW

13

own signal intelligence activities remain as decentralized as they now are." '\2 This bureaucratic ploy spurred the War Department to thoughts of reorganization. After consulting theater commanders in the Far East, the Army created a single unified signal s intelligence organization on 15 September 1945. The Army Security Agency established that day was to control "a ll signals intelligence and sec urity establishments, units, and personnel" and to function under direct control of the Military Intelligence Division. The U.S . Army thus attained the integrated COMINT organization it s intelligence chiefs had so long desired-thirteen days after the world war that had called it into being had finally come to an end . .\]

14

SIGNALS INTELLIGENCE

Notes Note: SRH c itation s refer to Spec ial Research Hi stories: declassified/sanitized cryptologic documents released to the National Archi ves by the National Security Agency. 1. SRH 134, " Expan sion of the Si g nal Inte lli gence Service from 1930-7 December 1941 ," pp. 3-28; Herbert O. Yardley, The American Black Chamber ( Indianapoli s: Bobbs-Merrill, 1931 ), pp. 369-372; Da v id Kahn , The Codebreakers: The Story of Secret Writing (New York: The Macmillan Company, 1967), p. 360. 2. SRH 29, "A Brief History of the Signal Intelligence Service," pp. 8- 13. 3. SRH 134, p. 19; Kahn , Th e Codebreakers , p. 24; SRH 106, "S pec ifi c Instructions for the Handling and Dissemination of Special Intelligence," pp. 1- 2. 4. SRH 125, "Certain Aspects of ' Magic' in the Cryptologic Background of the Various Official Investi gations into the Pearl Harbor Attack," p. 63. 5. SRH 141 , "Papers from the Personal Files of Alfred McCormack, Colonel, AUS, Spec ial Branch, G-2, Military Inte lli gence Divi sion , War Department General Staff," Part 2, pp. 7-8. 6. ULTRA alld the Army Air Forces in World War /I, edited by Diane Putney (Washington , D.C.: Offi ce of Air Fo rce Hi story, 1987), p. 71; SRH 349, "The Achievements of the Signal Security Agency in World War II," pp. 13- 14. 7. ULTRA and the Army Air Forces ill World War /I , p. 7 1; SRH 135, " Hi story of the 2d Signal Service Ballalion," p. 101. 8. SRH 349, pp. 9, 15; Ronald Lew in , The American MAGIC: Codes , Ciphers. and the Defeat of Japan (New York: Farrar, Straus, Giraux, 1982), pp. 143, 149. 9. SRH 141, Part 2, pp. 202-203; Part I, p. 29. 10. SRH 169, "Centralized Control of U.S. Army Si gnal Intelli gence in World War II ," pp. 14- 15. 11 . SRH 141 , Part 2, pp. 8- 13, 139,223; SRH 116, "Origins, Functions, and Problems of the Special Branch, MIS," p. 22. 12. SRH 349, p. 26. 13. FH. Hinsley et a\. , British IllIelligellce ill the Secolld World War (New York : Cambridge Uni versity Press, 198 1), Volume II , pp. 56-57. 14. SRH 391, " American Signal Inte lli gence in North Africa and Wes te rn

OVERVIEW

15

Europe," pp.6-7. 15. Ibid ., pp . 63, 94; " History of the Signal Intelligence Division, European Theater of Operations, United States AmlY" (SECRET), Part I, Chapter V, p. I, Army Cryptologic Records ; Hinsley. British Intel/igence, Volume 11, p. 742; SRH 124, "Operational Hi story of the 849th Signal Intelligence Service," pp. 10-12. 16. SRH 361, "History of the Signal Security Agency, Volume II: The General Cryptanalytic Problem ," p. 83; SRH 141 , Part 2, p. 223. 17. SRH 141 , Part 2, pp. 221-222. 18. SRH 349, pp. 18,26-28. 19. Hinsley, British Intel/igence, Volume II , pp. 57- 58; SRH 141 , Part 2, pp. 182- 183. 20. SRH 33, " Hi story of the Operations of Special Security Officers Attached to Field Commands." p. 16; SRH 61. "Allocation of Special Security Officers to Special Branch," p. 30. 21. ULTRA and the Army Air Forces, pp. 86-87; Ray S. Cline, The CIA Under Reagan, Blish. and Casey (Washington , D.C.: Acropolis Press, 1981), p. III. 22. SRH 141, Part 2, pp. 316-317. 23. SRH 349, p. 10. 24. Jbid., p. 14. 25. G.R . Thompson and Dixie R. Harri s, The Signal Corps: The Olltcome. United States Army in World War n (Washington, D.C.: Office of the Chief of Military History, 1966), p. 347; SRH 391, pp. 94-96, 104. Details on the organization of the corps-level signal service companies can be found in SRH 42, "Third Army Radio Intelligence History in Campaign of Westem Europe ." 26. SRH 169, pp. 54- 55,60-61; SRH 365, "History of Signal Intelligence Division of the Signal Officer, AFMIDPAC," pp. 15-23. 27. SRH 124, p. 24. 28. SRH 141 , Part 2, p. 220. 29. SRH 135, p. 12; SRH 349, p. 6; Thompson and Harris , The Signal Corps: The Ollleome, pp. 331, 338-339. 30. SRH 169, p. 59. 31. SRH 141 , Part 2, pp. 282-295; SRH 169, p. 58. 32. SRH 276, "Centralized Control of U.S. Army Signal Int e lligen ce Activities, 30 January 1939- 16 April 1945," p. 139: Memo, AC of S, G-2, subj: Army-Navy Communication Inte lli gence Board-Establi shment of (2 Mar 45) , Amly Cryptologic Records. 33. SRH 349. pp. 85- 88.

CHAPTER II The Road to Pearl Harbor These documents trace the evolution of U.S. Anny signals intelli gence in the period prior to United States involvement in World War II. SRH 29, A Brief History of the Signal Intelligence Service, was written by William F. Friedman, the organization's first head . It sketches the beginnings of Army signals intelligence in World War I, provides an unflattering account of Herbert O. Yardley 's Black Chamber organization, the legendary cryptanalytic unit secretly funded by the State and War Departments, and goes on to explain the reasons for tran s ferring the function from Military Intelligence to the Signal Corps in 1929. SRH 134, Expansion of the Signal Intelligence Service, was also written by Friedman. It charts the growing pains of this organization after 1930, and discusses how the efforts of Friedman and his team resulted in the breaking of the Japanese diplomatic machine cipher, codenamed PURPLE. The actual intercept of communications was performed by specially trained Signal Corps personnel. After 1938, these were brought together in a unique unit, the 2d Signal Service Company. The establishment of thi s unit is documented in an excerpt from SRH 135, History of the Second Signal Service Battalion , 1939-45. The success of the SIS in breaking PURPLE was shared with the Navy's corresponding cryptologic unit, OP-20-G, but was otherwise closely held. The intelligence product obtained from decrypted Japanese diplomatic communications was assigned the codename MAGIC and dissemi nated to only a small group of top officials. One of the most important messages contained in Japanese diplomatic traffic is reproduced From The Magic Backgrolllld of Pearl HarIJor, Voillme IV, Appelldlt (Washington, D.C.: Department of Defense, 1977), p. A-89 . However, the stringent security measures surrounding MAGIC impaired effective and sustained evalu-

18

SIGNALS INTELLIGENCE

ation of the material. The excerpt from SRH 115, U.S. Army Investi gation s Into the Handling of Certain Communications Prior to the Attack on Pearl Harbor, 1944- 1945 , shows the impact of this policy of tight compartmentation on the War Department military inte lli gence organization.

Frolll SRH 029. William F. Friedman, the first head of the Signals

Intelligence Service, was America's foremost cryptologist. Born in C:arist RlIssia to l ewish parellls, Friedman was brought to the United States by his family as a very yOllng child. He was originally trained as a geneticist, bllt his energies were soon diverted to the exotic field of cryptology when he was hired by the eccentric philanthropist George Fabyan to work at Fabyan's private "think-tank"--Riverbank Laboratories at Geneva, Illinois. Fabyan had a theory that Shakespeare's plays had actllally been wrillen by Sir Francis Bacon, and that this could be proved by deciphering a hidden cipher that Bacon had scattered throughout the plays. Fabyan was wrongas Friedman convincingly demonstrated many years latel; no such cipher existed--but his enthusiasms led Friedman onto a brilliant coreel: In 1942, Friedman wrote this succinct aCCOulll of the origins of the Army's World War II signals intelligence organization. As can be seen ji-om (he text, his billerness against his great colleague and rival, Herbert O. Yardley, continlied to run deep. A BRIEF HISTORY OF THE SIGNAL INTELLIGENCE SERVICE I. Prior to June 1917 no department of the Government conducted cryptanalytic activities whatsoever. From June 191 6 to abou t May 191 8 a considerab le amount of work along these lines was conducted purely as a patriotic enterprise at his own expense by Mr. George Fabyan, whose Riverbank Laboratories at Geneva, Illino is, organ ized a small group of fairly well-trained cryptana lysts to work upon such codes and ciphers as were forwarded by the War, Navy, State, and Justice Departments. The undersigned directed the cryptanal ytic operations and training at the Riverbank Laboratories from the time of the inception of thi s work until it s close thereat in 1919, except for a period of a year when he was 1st Lieut., MID, serving at GHQ-AEF in the German code so lving section. 2. In June 1917 the cryptanalytic activities of the War Department were initiated by Co lonel Van Deman, G-2, with the commissioning of H.O. Yardley as 1st Lieut., MID. Yardley, who had been a telegrapher at the State Department and had

THE ROAD TO PEARL HARBOR

19

taken some interest in c ryptography. was g ive n two c ivilian employees to ass ist him . The work grew rapidl y and by the a utumn of 19 17 the increased staff was organized as a section designated as MI -8, which was subdi vided into 6 subsecti ons with duties as indicated ve ry briefly below: ( I) Code and ciphe r solution subsection. - Thi s subsection was what would now be called the cryptanal ytic subsection. It was the la rgest of the subsections of MI -8 and perform ed the cryptanalytic work not onl y for the Wa r De partme nt hw also fo r all OIher Covernmem deparlmellfs, in eluding Navy, Slale, iuslice. and Ihe two cel1sorships--Cable and Postal , which were th en separate OI:~al1 iza{ions. (2) Code and cipher compilation subsection . - Despite the fact that under Army regulati ons the compilation and revi sion of codes was a function of the Chief Signal Officer, compilation activities under the Signal Corps were apparently in a moribund state. Infomlation having been received that the Germans possessed copies of the War Department Telegraph Code, M 1-8 deemed it advisable to establish a code compil ation subsection, and that subsecti on produced several codes such as Military Intelligence Codes No.5 and No.9, small pocket codes for secret agents, and the like. (3) Training subsection. - In addition to training its own personnel MI -8 trained the majority of the pe rsonnel sent overseas for cryptanal ytic duties with field forces. both AEF and Siberia. (4) Secret inks subsection. - A laborato ry was established for the preparati on of in visible ink s for use by o ur own agents. It also examined letters for secret ink writing and an average of over 2000 lette rs per week we re examined for the military and postal censorship from July I, 191 8 to February I, 1919. (5) Shorthand subsection . - This subsecti on was organized to handle texts in va rio us shorthand systems, especially German, which had to be deciphered. (6) Communications subsection. - This was established in MI -8 for hand ling messages to and from military attac hes and inte lli gence o ffi cers servi ng abroad. In a period of 9 months it sent and received abo ut 25,000 such messages, practically all in code. 3. At the height of its development, wh ich was reached in November 1918. MI-8 was. for those days, a rather large unit. consisting of 18 officers, 24 civ il ian cryptographers and cryptanalysts, and 109 typists and stenographe rs. The time had come for the establishment of a definite policy for the future. Now. the guiding heads of Military Intelligence at that time fully recognized the high importance and val ue of the services rendered by the cryptanalyti c bureau, because they had been in positions whe re the products of the daily act ivity of the bureau came directly to notice and they could not fail to note the influe nce and bearing which the work had. not only upon the military and naval but also upon the diplomatic, political, and economic phases of the conduct of the war. They therefore had praclical experience in the mailer and cou ld bring the weight of their position of influence and their actual experience to bear upon those in c harge of the purse stri ngs, wi th the

20

SIGNALS INTELLIGENCE

result that they were able to obtain funds sufficient to keep a fairly large organization intact for a year or two. An annual appropriation of $100,000 was recommended in a G-2 study dated May 16, 1919 [text withheld] to be used as follows: Rent, light, and heat .................................................................. .............. $3,900 Reference books .................... ............ ...... ....................................... 100 Personnel: Chief .. ...... ... ........................................................ .... 6,000 10 code and cipher experts @ $3,000 each ............ 30,000 15 code and cipher experts @ $2,000 each ............ 30,000 25 clerks @ $1.200 each ...................................... .. .30,000 Total ..................... $100,000 The item for " rent, light, and heat" is explainable when it is noted that the bureau was to be moved ftom Washington with a view to hiding its existence. Of the $100,000 recommended , the State Department was to provide $40,000 and $60,000 was to be provided for expenditure by the A.C. of S, G-2 on "confidential memoranda" against fund s pertaining to "Contingency Military Intelligence Division"-that is, by vouchers not subject to review by the Comptroller General. The paper containing the recommendations made by the A.C. of S, G-2 to the Chief of Staff was "OK-ed" and initialled by Acting Secretary of State Polk on May 17, 1919, and on May 19 it was approved by the Secretary of War over the s ignature of General March , Chief of Staff. The plan was put into effect, the bureau was installed in a private house at 22 East 38th Street, New York City, and all personnel together with existing records were moved thereto. 4. It is important to note that at that time no provision whatsoever was made for radio intercept as a source of raw mate rial for cryptanalysis. A few words on this subject may not be amiss. Radio intercept of fixed station material in the U.S. by the Army during our active participation in the last war was rather fragmentary. The work was, of course, a responsibility of the Signal Corps, but radio communications could hardly compete with cable and wire communications as a source of raw material because not only did the large governments use cable in preference to radio, but also radio as a means of communication betwecn widely separated fixed station s was then in its infancy. The Signal Corps did, however, have what was called "mobile tractor" units stationed on the southern border, [text withheld] and in late 1918 one large intercept station was established at Houlton , Maine, for the purpose of copying trans-A tlantic radio. The large intercept stations maintained by the Signal Corps in the AEF did furnish a large quantity of diplomatic traffic which was forwarded as raw material to 0-2 in Washington , but as soon as the war was over all these sources of intercept material were abandoned. [text withheld]

***

THE ROAD TO PEARL HARBOR

21

This was, of course, true to a certain extent but was not by any means the whole truth. At any rate by 1929 the flow of messages furnished G-2 had dwindled to a mere trickle. In about the year 1926 the intercept stations of the Coast Guard submitted to me infornlally a few radio intercepts of what appeared to be Japanese code messages, but these were only sporadic occasional intercepts of transmissions in the Far East. They were usually fragmentary and afforded very poor material. Such as they were, however, these occasional messages were turned over to G-2. There were, therefore, clear indications that if a serious attempt were made to obtain this material better copy could be obtained. However, no pressure was brought on the Signal Corps to set up a regular intercept service. and aside from a few relatively weak efforts to establish an intercept station in the Philippine Department there was no real intercept service until after 1929, when responsibility for cryptanalysis was transferred from G-2 to the Signal Corps. How this came to pass will soon be detailed. 5. As indicated Par. 3 above, the $100,000 appropriation set up in 1919 took care of the bureau for the FY 1920, that is from July I, 1919 to June 30, 1920, but when in June 1920 it came time to set up the budget for FY 1921, the purse strings were already beginning to be pulled tighter. Many of the "old-timers" in G-2 had gone to other assignments; those remaining and the newcomers in G-2 apparently had neither the background nor the influence to press the matter. The appropriation was at once cut in half, that is, to $50,000, of which the State Department share still continued to be $40,000. The theory was, evidently, that since the work done by the bureau was primarily, if not solely. for and of interest to the State Depanment, all or nearly all of the funds should be provided by that department. The War Department overlooked some very important points in the situationpoints which will be brought up and emphasized later in this summary. In order not to break the continuity of the history at this point. it will be stated merely that year by year the funds provided for the maintenance and operation of the bureau became more and more constricted until by the autumn of 1929 the following tabulation, based upon a letter dated July 17, 1929 from Major O.S. Albright, G-2 to the Chief Signal Officer (General Gibbs), shows how the bureau had been pernlitted to deteriorate: Rent ......................................................... ..... .. ................................... $3 ,000 Books, postage, travel. transportation, misc . ... ................. .................. 2,370 Personnel: I Chief (Yardley) ............................... ........... .................. 7,500 I Code and cipher expert ................................................ 3,660 I translator (Jap) ..................................................... ...... ... 3,750 I secretary ..................................... ... ............................... I ,800 I clerk-typist ................................................................ ... 1,600 I clerk-typist ... ... ............................................................. I ,320 Total .. ........................ $25,000

22

SIGNALS INTELLIGENCE

In the foregoing tOlai appropri ation o f $25 ,000 the State Department furni shed $ 15,000, the War Department $ 10,000. The ac ti vities of the bureau had by thi s time become so reduced lhat it was sending in onl y occasional translations of a few Japanese [text withheld] diplomatic messages. No research whatsoever was conducted in cryptanalysis; the re we re no training ac ti vities, no inte rcept , no code compilation, no secret ink wo rk. The pe rsonnel consisted o f six persons all told and 37 1/2% of the total payroll went to one man, who had little interest other than to continue as long as poss ible to maintain him self in the sinecure into whi ch he had been pe rmitted to establish himself. 6. In the s umm e r of 1929 Maj o r O .S . Albri g ht , S ig na l Corp s, had bee n ass igned to G-2 to serve on the Staff o f the A.C. o f S., G-2 to supervise and coordi nate such o f the cryptographic and cryptanalytic acti vities of the War De partme nt as re mained. Afte r a careful stud y o f the situ ation and a n a ppraisal o f how the ex isting cryptanalytic bureau was and was not serving the functions fo r which it had been intended, Major Albright came to the conclusion that the e ntire picture was wrong. He felt that the product (" bulletin" ) whi ch the bureau was turning out every few d ays was indeed o f prim ary int e rest for its own sake to the St ate Department a nd while the War Department had onl y a secondary inte rest in the " bulle tin" fo r the information it gave, the primary interest o f the War Department in cryptanalytic studies in peace time was that it was inte nded to prov ide a means for training of personnel for immedia te war- time effec ti veness. M aj or Albright found that not onl y was there very little if any tra ining bei ng conducted but also that all the pe rsons in the bureau. except for one cle rk receiving the least pay. we re "getting along in years"-their pote nti al use fulness for poss ible war time se rvice prac tica ll y nil. Moreo ve r, the bureau was now hidde n away in a public o ffice buil d ing in Ne w Yo rk (unde r cover o f the "Code Compil ati o n Company" fo r alleged purposes of securit y) and fa r away from direct supe rvision of an ybody connected with the War Departme nt or o f G-2, so that nobody knew what was going on. how the o ffice was administe red. etc. Yard ley devoted most of his time to two or three pri vate ente rprises (commercial code compilation, rea l estate broke rage, consult ant in code matte rs to comme rcial firm s) and he was hav ing a "field day" at Governme nt expense. There were, in additi on. several othe r weighty fac tors which moti va ted Major Al bright in pre paring a G-2 stud y [tex t withheld [ recomme nding th at the bureau be take n out o f G-2 and it s fu nctions transfe rred to the Signa l Corps, chief among which was the desirabil ity. if not necessity, of placing all cryptogra phic and cryptanal ytic work of the War Deparme nt unde r one agency, rathe r th an distributing it among three (Th e A djutant General, for printing, storage, iss ue, a nd acco untin g o f codes; Th e C hi e f Si g nal Offi ce r, for compiling cod es and ciphe rs; Military Intelli gence for soluti on o f codes and ciphe rs). A memo randum on the same subject was pre pared by Li e u!. Col. W.K. Wil son of the War Plans and Training Section of G-2, and is also attac hed he reto [text withheld] The reasons g ive n in Major Albright 's stud y and in Col. Wil son's memorandum we re appare nt-

THE ROAD TO PEARL HARBOR

23

Iy deemed valid by the Chief of Staff, for Major Albright's recomme ndations were a pproved in April 1929 and steps we re soon initiated by G-2 and the Chief Signal Officer to put the m into e ffect. The recommendations carried with them merel y the wording of a changes to be made in AR 105-5, specifying the duties of the Chief Signal Officer, these duties being enlarged to include the printing, storage, distribution , and accounting of codes and "in time of war the inte rception of enemy radio and wire traffic, the goniometric location of enemy radio stations, the solution of intercepted enemy code and cipher messages, and laboratory arrangements for the employment and detection of secret inks." 7. However, before anything could be done actually 10 tra nsfe r the ac tivit y a new a nd ve ry di sturbing factor e nte red into the picture. In March 1929 a new administration took o ffi ce, in which Mr. Stimson became Secretary of State. For a few weeks no " bulletins" from the cryptanaly tic bureau in New York were g iven him. the intention being to "go slow" until he had become sufficiently well oriented in the duties of his offi ce to warra nt bringing to hi s attention the highl y secret activities engaged in by War and State De pal1me nt s by means of funds provided in large part by the latter Departme nt. Early in May 1929, however, the time was deemed ripe for thi s measure, but it was with some trepidation that a few translations of Japanese code messages were placed on Mr. Stimson's desk. Hi s reaction was vio le nt and hi s action drastic. Upon learning how the material was obtained. he characterized the activity as being hi ghl y unethi cal and declared that it wou ld cease immediately, so far as the State Departme nt was concerned. To put teeth into hi s decision he gave instructions that the necessary funds of the State Departme nt would be withdrawn at ollce . It was only after considerable pressure by the A. C. of S., G-2 that he was dissuaded from thi s course, whi ch might have had serious consequcnces by sudden ly throwing out of e mployment the six people concerned, at a time of severe economic depression. For these worke rs had only special training in a field wholely useless to commercial, industrial . shipping or banking firm s. or to other governme nt departments, or to educat ional institutions. An arrangement was the refore made to c lose the office immediately so far as ac ti ve wo rk was concerned but to keep the personnel on the payroll for the time necessary to wind up affairs, and get the files in shape ready to turn over to the Signal Corps. This took a coupl e o f months. and at the end o f June 1929 the employees we re g ive n three mont hs' pay illlld"{lIICe in a lump sum , to tide them over the period in which they might be jobless. Since they had been paid out of "confidential funds" the y had no civi l service status and no retirement benefits; moreo ver. they we re inelig ible for transfer to other Government positions. The danger was, o f course, that their di ssatisfaction with what must have appeared to them as hi ghhanded. arbitrary action on the part of a new official, and that their helplessness in the serious personal si tuation created for them by hi s drastic action might lead them to indi scretions which might prove most embarrassi ng to the Government and have serio us consequences upon national defense. It turned out that whatever the ir private feelings, all the dis-

24

SIGNALS INTELLIGENCE

charged personnel, except the chief beneficiary of the old regime, remained loyal and did the best they could to find jobs. 8. In October 1929 I was sent to New York to take over the boxed up records and files and to oversee their transportation to Washington. The cryptanalyt ic activities, research, and training now being under the Chier Signal Officer, steps were taken to reorganize the bureau and its work. The funds available were, of course, ve ry slim-on ly what remained of the War Department's contribution of $10,000 for the FY 1930 was available, because the remainder of the State Department's share of $15,000, as indicated above, had already been withdrawn by the State Department. An offer of employment was made to Mrs. Wilson, the Japanese expert wi th Yardley, but she was unable to accept, since it involved moving to Washington and she had a husband and child in New York. Another employee, Mr. Victor Weisskopf, had a business in New York and refused to move to Washington. The female clerical emp loyees were deemed unsuitable for our purposes and, moreover, having no Civi l Service status they could not be taken on by transfer. An offer of temporary employment was made to Yardley but he refused the tender. Instead, he proceeded secretly to prepare a book which first appeared in the form of articles in the the Saturday Evening Post and which in much expanded form later appeared under the title The Americall Black Cham her. The book and articles were highly sensational and made damaging disclosures concerning the most secret activities ever conducted by the Government. Before the appearance of the articles and book, however, he had taken certain steps to protect himself from possible prosecution for his disclosures, among which was to resign his commission as major in the Military Intelligence Reserve. Of course, had the authorities understood the real purpose of his resignation they might have prevented it so as to retain some hold on him. But being in ignorance of the real motives and deeming it just an act of pique, the resignation was accepted. The unfortunate consequences attendant upon the publication of the book need no elaboration herein. Suffice it to say that [text withheld] our precarious relations with Japan were brought to a boiling point when about 30,000 copies or the Japanese translation of The Americall Black Chamber were sold in Tokyo in a period of less than a month (perhaps the book was subs idized by the Japanese Government itself). The bad odor into wh ich all cryptana lysts and cryptanalytic activities fell, as a result of the difficulties which the publicity given the matter by Yardley's disclosures occas ioned high government officials, had a bad effect upon the attempted reorganization of the cryptana lytic bureau by the Chief Signal Ofricer. FlIllds were hard ta get, and State Department support was lack ing, if not in the other direction altogether. The most serious consequences of Yardley's disclosures, however, came 10 years later and their efrects can hardly be estimated. I refer here to the jolt which his book gave the Japanese cryptographers, lead ing them out of their blissful ignorance and caus ing them to develop really complex methods which are now giving us many difficulties. The same is true probably as regards the Gennan and Italian cryptographers-their

THE ROAD TO PEARL HARBOR

25

education has been entirely at Uncle Sam's expense and the final consequences of Yardley's work can not yet be foreseen. They may well tum out to be disastrous. 9. Howeve r, the Chief Signal Officer proceeded, e nergeti ca ll y as possible unde r the circ umstances, to carry out the mi ss ion assigned him . The code and c ipher solving section was placed unde r the War Plans and Training Division s ince the compilation section was already the re. A rather detailed directive which was prepared by G-2 and approved by the Secretary of War (Exhibit 4), became the g uidin g plan of the reorga nized serv ice, which was no w nam ed the Signal Inte lligence Service. Its personnel consisting of myself and one or two clerks soon was augmented by a half dozen more employees. Training literature and courses were pre pared and put into good usage at once. A great deal was done in expanding our cryptographic work also, by preparing reserve editions of existing codes. co mpilin g new codes and ci phers, developing c iph e r apparatus, and so on. C ryptanal ytic work was put on a firm basis of research and training, with emphasis on the laHer, jor there ex isted no intercept service and the raw mate rial could not be obtained [tex t wit hhe ld] He nce an intercept service now was organized and grew very slow ly. All phases o f signal intelligence were unified under one service and taken under sllldy and action. Moreover, cooperati on with the Na vy in the work was also initiated. How the activity has ex panded since the n requires no comment at this time. However, a few wo rds about relations with the Navy are pertinent.

10. Cryptanal ytic activities in our Navy Departme nt were practically non-ex istent until after the close of the last war, during which, as was noted above, whateve r problems they had in c ryptanalys is were refe rred to MI -S. But in 192 1 the Nav y, recognizing the important role which cry ptanalysis was bound to play in the future , began building up a large unit in the Navy Depa rtme nt , with eche lons afloat. Whereas the Army placed emphas is upon c iv ilian trainin g, the Navy placed emphasis upon officer trai ning; and for each dollar the Army was able to obtain for cryp tanalytic and cryptographic work the Navy was able to obtain three to fi ve dol lars, until by 1939. as far as conce rned numbers of offi cers and civilian personne l e ngaged in these activities. amount of equipment on hand. and funds available for research, the Navy had conside rabl y outstripped the Arnly. Howevcr, it may be said, with some justifiable pride pe rhaps, that while they werc ahead of us in quantit y we were a head in qualit y, for all the important developments in both the cryptographic and cryp tanal yt ic fi elds must be credited to Army personnel. At first , cooperation betwee n the two services was inte rnliHe nt and at times very indifTe rent-the usual mUlllal suspicions and jealousies pervaded our relat ionships. But, happil y, for the past three to five ycars coope rati on has been muc h more who lehearted, with the result that it may now be said without reserve that , as regards thei r cryptograph ic and cryptanal ytic acti vities. cooperatio n between the Army and Navy is so close as to be the same as though they were under one head. This, of course. is as it should be and must be in order to gain the desired result from such acti vities.

26

SIGNALS INTELLIGENCE

II. It would be of utmost value to the winning of this war if the Government were now in a position to read the codes and ciphers of all the foreign powers whose actions and probable intentions are of interest and importance in our prosecution of the war. It could have been in this fortunale posilion had it given to cryptanalytic studies the attention which it deserves during peace lime alld had provided funds for their contilluity all a scale sufficient for the purpose for which they are il11ended. The matter can be summarized very succinctly in this statement: Actual or physical warfare is intennittenr, but mental, that is, cryptanalytic warfare is continuous . ...

*** William F. Friedman, Head Cryptanal YSl.

June 29, 1942

From SRH 134. This official history, wrirren by Friedman in 1945, traces the history of the SIS ji"om its modest beginnings through its triumph in decrypting the Japanese diplomatic machine cipher system. It should be noted that Friedman had no doubt that the activities in which the Signal 111Ielligence Service was engaged were quite illegal.

EXPANSION OF THE SIGNAL INTELLIGENCE SERVICE FROM 1930T07 DECEMBER 1941 As a result of War Department General Staff action it was directed by the Secretary of War, on 5 April 1929, "That the Signal Corps be charged with the duties pertaining to the solution of enemy codes and ciphers and the detection of secret inks in War. in addition to those duties with which they are now charged pertaining to the compilation of codes and ciphers required by the Army and to the interception of enemy radio and wire traffic in war."

*** ... the Signal Intelligence Service was to be organized primarily on the basis of training for war, rather than for active operation for enemy interception and solulion of the communications of foreign governments or amlies. However, it was also

THE ROAD TO PEARL HARBOR

27

stated that if infornlation of present or immediate value to G-2 should be intercepted and solved, such information would be acceptable to G-2, but it was to be clearly recognized and regarded as a by-product of the training work, and not as the fruit of the functioning of the Signal Intelligence Service in peacetime. But in addition to thi s foregoing reason for the emphasis on training. there was another important reason. Because of the rigidit y oj' our laws regarding interception of communications, the activities of the Signal Intelligence Service had to partake of the nature of a clandestine activity, since it was illegal to intercept messages, and the penalties for divulging even the existence of a message were severe. The law did not exempt anyone from it s penalti es, and therefore, the War De partment and all the people connected with the work of interception and solution of the communications of foreign go vernments realized the necess ity for keeping the activity secret. By emphasizing the training aspects of its activities, it was thought that the illegal nature of some of the work of the Signal Intelligence Service (interception and so lution of foreign communications) could be more eas ily defended if word of these activities should " leak out." The foregoing reason for secrecy was, of course, 10 be added 10 another important reason for keeping the activity secret - namely, the requirement to keep from foreign governments the ex tent of our success with their communications, and even the fact that we were successful in reading any of such communications. Under such circumstances. the refore. it was only to be expected that the normal channels for Obtaining of funds for such an activit y would be made somewhat difficult and they were. Up to thi s time, it must be understood, the War Department support for the so lution activities came from Military Intelligence Division Contingent Funds, which are not subject to review by the Controller General. For one reason or another, none of the six persons who had becn working in the old MI-8 Section when the Chief Signal Officer took over the solution activities from G-2 could be added to the staff of the Signal Intelligence Service, which then consisted of one cryptanalyst, (P-S, $4,600) and one clerk-typist (CAF-3, $ 1,620). A whole new organization had therefore 10 be built -up with the very scanty fund s ($6,666.68) that remained from the original War Department contribution of $ 10,000 for the rest of the fiscal year 1930. Therefore, to begin with, the real establishment of a permanent Signal Intelligence Service had to await the availability. to the Chief Signal Officer. of funds specifically for the purpose; and the n, of course. the long period of training required to produce c ryptographic and c ryptanalyti c skill was another faclOr that resulted in what may appear to be a rather slow development. in addition to the usual lack of fund s. However. the hiring of new personnel did not have to await the availability of funds , for until the Chief Signal Officer was able to include in hi s own estimates the specific fund s necessary for the support of the Signal Intelligence Service, he could draw by allotment upon the funds available to the Assistant Chief of Staff, G-2. These, howe ve r, amounted to only $ 10,000 per annum. The sum of $6,666.68

28

SIGNALS INTELLIGENCE

was allotted by the Ass istant Chief of Staff, G-2 , to the Chief Signal Officer For the payment of personnel e ngaged in code and cipher work on 16 December 1929 for the balance of the fiscal year 1929. On 4 January 1930 the Secretary of War was requested by the Chief Signal Office r to authorize the employment of four junior cryptanalysts (P- I, $2,000 per annum). Since the G-2 funds ($10,000 pe r annum) were available to the Chief Si gnal Officer the FY 1930 and 193 1, thi s request was approved quite promptly; three persons were obtained from Civil Service rosters of eligibles in sc ientific pursuits (j unior mathe mati cian s, e tc .) a nd appointed . An additional person was obtained by Schedule " A" or excepted appointment for the position of Japanese translator (S P-5, $ 1,800 per annum) because no Civil Service roster of eligibles was available for se lection. On 2 September 1930 an additional pe rson (S P-5, $ 1,800 per annum) was appointed, makin g th e original staff of th e Signal Intelligence Service a total of seven persons by the end of 1930. Except for one or two changes in personnel and the addition of one or two clerk-typists, the technical staff was not augmented until Jul y 1938. In 1936 and 1937 a cryptanalyst was sent to each of the Panama Canal and Hawaiian Departme nt s to work as advanced echelon s o f the Signal Intelligence Service, and these two pe rsons' positions were filled by two additional candidates selected From Civil Service rosters. However, by the end of 1938 both cryptanal ysts who had been sent to overseas stations were back in Washington, for the experiment was not a success, a nd we badly needed them here. What could one lone cryptanalyst do in Panama or Hawaii? The years 193 1 to 1936 were particularly trying ones - promotions were out of the question, and further expansion was great ly handicapped by the program of economy imposed upon the War Departme nt during the depression. Several examples may be cited to illu strate our attempts to expand the Signal Intelligence Service over the yea rs 1931 - 1938 and the more or less fruitl ess results the reof. On 14 October 1931 the Chief Signal Office r requested approval of a project to include four additional positions in the Signal Intelligence Se rvice, in volvi ng an increase of $7,240 in hi s estimates for sa laries for the fi scal yea r 1934. As inclosures, the Ch ief Signal Officer submitted schedules sho wing the number of people engaged in Signal Inte lligence work in fiscal yea rs 1928- 193 1 and projecting the numbers estimated by 1934. The request for the four additional positions was " not favorably considered at this time" by the Adjutant General. The reason g iven was the necess ity for holding estimates to a minimum for the fiscal years 1934 : " It is not contemplated outhorizing new projects unless c urrent or prospecti ve funds are available. "

Again in 1932 an attempt was made to attain urgently needed personnel, with no favorable results. It was stated that the need for expanding the code producti on program was urgent. In addition to the code compilation duties of th e sec-

tion, there were tho se of cond uctin g research in so lution as we ll as certain administrative duti es in connec ti on with corre spondence courses, reserve officer

THE ROAD TO PEARL HARBOR

29

instructional training courses . and the regular administrative activities concerned in th e lise, distribution and improvement in code and cipher communi cation by

the Army, the developme nt of c ipher apparatus, methods of secre t inte rcommuni cation between the Army and Navy. the developme nt of radio inte rcept and radio direction-finding me th ods, work in sec re t inks. e tc. The ass istance of G -2 was so ug ht at the time . since it was thou g ht that in view of the inte rest regis tered in the work of the Signal Inte llige nce Section o n the part of the Assistant Chief of Staff, G-2, and the Chief of Staff, it would be the psychological moment to take the mailer up once more in the hope that means mig ht be found to add to th e pe rso nnel of the secti o n , and thus enable it to ca rryo n in an effecti ve manner. Whether or not the suppo rt of the Assistant C hi e f of Staff. G-2 , was obtained , the available record does no t di sclose; ho weve r. even if thi s s uppo rt was forthcoming. the allempt re mained fruitle ss. In 1936. the C hief of the Signal Inte lli ge nce Service recomme nded that the time had come to organize the Signal Inte lli gc nce activities if the Chief Signal Officer upon a more extensive basis, " in order that personne l for e ffi c ient operation may be available when the situation will require th eir servi ces. " II was al so

cons ide red essential to provice opportunities for advancement for the pe rsonnel already employed, in orde r that a restri c ted fi e ld mi g ht be allractive to them . Otherwise. the Signal Int el li ge nce Service wo uld "become m e re ly a training gro und for o ther departments." A five-year expansion program was recommended. which wou ld increase the to tal personnel to 2 I by 1942. with a to tal budget of $54,660. Various obstacles impeded the immediate approval of thi s plan . In the first place, in order to be able to keep our trained personnel. long-overdue promotions were very necessary: but any proposal fo r an inc rease in the sala ries of the pe rsonne l in the Signal Inte lli ge nce Service was held to be objectionable. (For several years. ow ing principa ll y to measures of economy imposed by the Preside nt and Congress, it had not been possible to provide administrative promoti ons for any Sig nal Corps employees. In addition. the President had directed that no promotions were to be included in the 1937 budgel.) A second obstacle was that the personnel a nd equipme nt assigned to the Signal Inte lligence Service fully occupied it s available space, and because of the cri ti ca l shortage or office and storage spacc in the District of Co lumbi a. no additional space was like ly to be secured for the ex pans ion of that organi zation . It was sugges ted that add it io ns to the ex isting force wou ld overc rowd the area and be detrimental to health and comfort as we ll as impair the performance of th eir

duties. This request for pe rsonne l was not approved and the four add iti o nal posit io ns had to be deleted from th e esti mates. One of the co ntributo ry factors which brought about thi s de let io n was the lack of s uppo rt frolll representatives of G-2. Assurance had been obta ined from G-2 on advancc that it wo uld coopera te in

SIGNALS INTELLIGENCE

30

defense of the item , but its representatives "failed to appear at the hearing when held. " Note the following extract from a very interesting memorandum dated 18 February 1936, to the Chief Signal Officer, again requesting an increase in personnel for the Signal Intelligence Service: " 9. Under the date of March 20 , 1935, the Chief Signal Officer requested authority to present in connection with the 1937 estimates, a requirement for two (2) assistant cryptograp hic clerks at $1,620 per annum, and two (2) junior cryptanalysts at $2,000 per annum. The War Department approved the inclusion of thi s additional personnel in the Signal Corps estimates. The limiting figure, des ignated by the War Department under date of Augu st 5, 1935, for Projec t #1, "Personnel", of the Signal Corps estimates to be s ubmitted to the Bureau of the Budget, was reduced to such a point as to necessitate the deletion of these four positions from fi scal year 1937 estimates. One of the contributory factors in the necess ity for this deletion was, no doubt, the lack of support by representative s from the Ass istant Chief of Staff, G-2, whose assurance of cooperation in defense of thi s attempt had been obtained in advance, but who failed to appear at the hearing when held."

At that time it was the opinion of Major W. S. Rumbaugh, Officer in Charge of the War Plans and Training Divi sion, of which the Si gnal Intelligence Section was a part , that such "a serious shortage of trained personnel exists in the Si gnal Intelligence Servi ce." that it could not " fully perform its peace-time mi ss ion ."If thi s shortage should " be allowed to continue, no Signal Intelligence Service worthy of the name will be available during the early phase of an emergency when the most valuable results should be ex pected from thi s agency." The memorandum of 18 February 1936 recommended that if funds in the amount of $7,240 acc rue by sav ings, under Project #1, fiscal year 1937, that these savi ngs be used for the employment of two Ass istant cryptographic clerks at $ 1,620, and two junior crypt analysts at $2,000, to be obtained and trained in thi s office as soon as possible ; alternatively, that if no such savings can be foreseen, fund s be diverted from other projects and allocated to Project # I, fiscal year 1937, for the foregoin g purpose. But , as already noted , the attempt was fruitle ss. It may be stated that about thi s same time, in May 1936, whereas the Signal Inte lli gence Service had a total of some seven or eight people, the Navy had some forty, and even the small unit at the United States Coast Guard had seven people. [text withheld] All attempts to increase the complement of seven people compri sing the staff of the Signal Intelligence Service were more or less fruitle ss

THE ROAD TO PEARL HARBOR

31

until the fiscal yea r 1938, when four more positions we re finall y authorized. No doubt one of the factors which led to the final approval of the repeated requests for expansion was the recognition on the part of the hig h authorities of the War Department of the importance of the work being done. A program of code compilation was being vigo ro usly pursued unde r severe handicaps; new codes and ciphers were being prepared, printed and placed in storage with two, and sometim es three or more ed it ions in reserve to meet all contingencies. ( It is necessary to note that in 1933 the duties formerly assigned to the Adjutant Gene ral in con-

nection with the printing. distribution, and accounting for codes, we re transferred to the Chief Si gnal Officer,) thus integrating all cryptogra phi c work in the Army within one organization-the Signal Intelligence Service under the Chi ef Signal Officer.) Very important progress in research and development of new cryptographic appa ratus a nd devices was being made: and a hig h-speed , e lectri ca l, c ryp tographic machine of hi ghest sec urit y had been invented and developed by o ur chief c ryptanal ys t- thi s mac hine being far in advance of anything that had hithert o been known. M oreover, thi s machine was going into practical usage for secret interc omm uni ca ti on among War Department and th e Comm and ers of

overseas departments. An intercept service was being organised to provide the raw material for the cryptanalytic training ac tivities. the first stat ion to be established being that at Fort Monmouth. New Jersey, with additional stations being contempl ated in th e ove rseas departments. On the crypt analy ti c side much progress had already been made, and by 1934 the training program had reached the stage where actual so lutions of c urrent Japanese diplomati c messages were being achieved dail y. It is, however. a com me ntary on the peculiar situ ation whi ch ex isted to state that so fearful was the C hief Signal Officer th at news of the "i llega l" activities of the Signal Inte lli gence Service might leak out. that these so luti ons were not issued as "b ull etins" or distributed to anybody. Occasiona lly, one wo uld be sho wn to the Chief Signal Officer. and it is possible that he wo uld show a translati on to the Assistant Chief of Staff. G-2, but it is doubtful if dissemination went any farther than that. Whe n important vis it ors from the War Department we re brought by the C hief Signal Officer to the quarters occupied by the Signal Intelligence Service for inspection or other officia l business, it was important to emphasize thai such so luti on work as was being accomplished was onl y a by-product and part of the tra ining being conducted, and that no use was made of the product. It was not until 2 March 1935 that "publication" of the bulletin was started , and G-2 began receiving on a more or less regu lar schedu le the translations of Japa nese diplomatic messages resulting from the " training." What distribution o r dissemination G-2 made is. o f course, not known to me, except by indirect ev idence . It was not until 1939 that any considerable increase in funds for the Signal Intell igence Service was provided, as is shown by the following letter:

32

SIGNALS INTELLIGENCE

September 27 1939 "SUBJECT: Signal Intelligence Service "TO: The Chief Signal Officer ., I. You are authorized to employ the following designated additiona l civilian personnel from October I. 1939, at the annual salaries indicated. I Principal Cryptanalyst @ $5,600 ......... .......................................... $5,600 2 Associate C ryptanalysts @ $3,200 ................................................. 6.400 2 Research Analysts@ $3,200 .... ............................ .......................... 6.400 2 Assistant Cryptanalysts@ $2,600 ....... .. .. ...... ........ ........................ ' 5,200 2 Junior Cryptanal ysts @ $2,000 ...................................................... .4.000 2 Cryptanalytic Aides @ $ 1,800 .................................... .................... 3,600 4 Principal Translators @ $2.300 ...................................................... 9,200 I Principal Clerk @ $2.300 .............................................................. .2,300 I Principal Stenographer @ $ 1,800 ................................................... 1,800 3 Cryptographic Clerks @ $ 1,800 ..................................................... 5.400 I Senior Tabulating MachineOperator @ $ 1,800 .... .. ...... ................. 1,800 3 Assistant Cryptographi cClerk @ $ 1,620 ...................................... .4,860 2 Tabulating Machine Operators @ 1,620 ......................................... 3.240 "2. It is desired that you establi sh and operate a mon itoring stat ion at Fort Hunt , Virginia, and purchase such of the following listed eq uipment as may be necessary and procurable within the funds allotted: 10 Radi o Sets, SCR 243 (or equivalent) @$260 ................... .......... $2,600 29 Radio Sets. SCR 244 (or equi va lent) @$ 195 .................. ............. 5,655 18 Recorders, BC 17 @ $ 1,300 ...................... ......................... ....... 23.400 12 Recording Equipment Tape @ $2.000 ........................................ 24,000 9 Diversity Antennae @ $3,000 .................................................... .. 27,000 14 Diversity Receiving Equipment @ $ 1.400 .. ................ ........ ....... 19.600 TOTAL ................. $ 102,255 " 3. The authorized st re ngt h of th e Second Signal Service Company is increased by the addition of th e following li sted personnel. The a utho ri zed strengths of Signal units from wh ich these men are transferred are red uced by the correspond ing grades and specialist ra tings . The allotment of enlisted men 10 the Signa l Corps is increased by twenty-six privates (grade 7). No increase in grades above grade 7 or in special ist ratings will be made until such time as additional grades and ratings become avai lable (probably Nov. 8, 1939). I Master Sergeant I Technical Sergeant I Staff Sergeant I Private 1st Class, spec. 1st ciass 6 Private 1st Class, spec. 2d ciass

THE ROAD TO PEARL HARBOR

33

6 Private 1st Class, spec. 3d class 6 Private 1st Class, spec. 4th class 4 PriVaie 1st Class, spec. 5th class "4. II is desired that you submit the name of and other perlinem data concerning a Regular Army Officer to be ass igned as assistant to the officer in charge, Signal Imelligence Service, and the names of and othe r perlinelll data concerning nine Reserve Officers for active dut y under the provisions of the Thomason Ac!. " 5. Funds in the amoun! of $41 ,050 from the Administrative Reserve Signal Service of the Army and in the amoun! of $50.000 from the reserve now in an unallOlled status, subject only to direction of the Secretary of War, will be released to you for the salaries of these civilian employees, for the purchase of additional equipment, and for the expenses incident to operation of the Signal intelligence Service. The balance necessary for the purchase of eq uipme nt will not be available for obligation ulllil the total defici ency of $ 175,405, required for the expansion of the Signal Intelligence Service has been covered by Cong ressional appropriations. "6. You will make necessary arrangements in collabo ralion with A.C. of S. G-2 to obtain C ivil Service exemptions required. " By Order o f the Secretary of War: /s/ WILLIAM w. DICK Adjutant Genera l" The total numbe r or additional personne l for the Signal Intelligence Service in Washington authorized by the foregoing lelle r was twe nty-s ix ; with the original staff of nineteen people, the Signal Inte lli gence Service starf was 10 be fOrly -fi ve persons. Steps were immedialely taken to hire the new personel, have them cleared, and to start them in training - a rather lengthy process. Special di spensalion had to be obtained rrom the Civil Service Commission to hire these trainees outside the C ivi l Service regulations, because no rosters of e ligibles in cryptographic or cryptanal ytic wo rk existed. In order to obtain the support necessary for such an unusual procedure the Secretary of War requested approval by the President on 26 October 1939 to have the C ivil Service regulations suspended so as to pe rmit the emp loyment by the Signal Inte lligence Service of not more than twellly-six persons on a temporary basi s until 30 June 1940. Through the interventi on or the White House, the approval requested was obtained. The un favourab le way in which the war in Eu rope was progressing so rar as concerned Britain, led to desires to expand th e Signal Intelligence Service even

more, and as fas t as possible. The urgency for cryptographic production to meet the needs of a large United Stales Army in case the United States became a belligerent was apparem; the output of the c ryptanalyti c starr was impressive enough 10 G-2 to warran t furl her increase of that staff. By September 1940 the Signal Intelligence Serv ice. without the help of any othe r staff, had succeeded in a most momentous

34

SIGNALS INTELLIGENCE

achieveme nt. Itex t withheld) - the so lut ion of the Japanese diplomat ic mac hine designated by them as the "8 Mac hine" and by us as the ·'Purple." Texts o f highest strategic and diplomatic import we re be ing furni shed G-2 dail y, the value of which was recogni zed by the Assistant Chief of Staff, G-2, the Chief of Staff, the Secretary of War, and the President. The conduct of our diffi cult negotiations with the Japanese in regard to the situation in the Pac ific was no doubt being guided from day 10 day by means of the soluti ons fumi shed the War Departme nt , the St ate Department , and the President. On 16 October 1940 the Chief Signal Officer requested of the Assistant Chief of Staff, G-2, authori zation fo r two furth e r ex pan sions o f the S igna l Inte lli gence Serv ice, one of which, to begin I January 194 1, would involve appointment of fortyeight additional civilian employees at annual salaries totaling $85,580, half of which ($42.790) wo uld have to be provided during the fisca l year 194 1, then already in progress. No funds for this purpose were available to the Chief Signal Offi cer, but this expansion was approved by the Ac ting Assistant Chief of Staff, G-2, with the concurrence of the Assistant Chief of Staff, War Plans Division, on the very next day. The second request, made on 16 October 1940, was for expansion involving employment of forty- two other employees, who we re to begin work on I Jul y 1941. T his pl an was approved in princi ple by Brigadier Gene ral Shennan Miles, Ass istant Chief of Staff, G-2, on 17 October 1940, but it s carrying out was postponed until the time fo r the prepa rati on of the budget for fiscal year 1942. The parag raph is of inte rest: "3. The complete ex pansion o f the Signal Intelligence Se rvice to place it on a wartime basis beg inning Jul y I, 194 1, as recomme nded in pa ragraph Ic and inclosure of the bas ic lelle r, is not essenti al at the present tim e. The recolllmendation, however, is

for funds to e nable this expansion to be begun on Jul y I, 194 1. The uncertainty o f the future and the necessity in any contempl ated ex pansion of th is service to make provision for a time lag incident to selecti on and training of new personne l make it necessary 10 provide for an increase in pe rsonnel fa r in adva nce. It is recommended, the refore, that fund s be made ava ilable in the fiscal yea r 1942 budget to prov ide fo r total ex pansion of the Signal Intelli gence Service. If condi tions develop pri or to I Jul y 194 1, whi c h make thi s addit ional expans io n unnecessary, the funds need not be expended." The letter from the Chief Signal Officer to the Assistant Chief of Staff, G-2, on 16 October 1940, contempl ated an ex panded Signal Inte lligence Serv ice fo r fisca l yea r 1942 amounting 10 127 pe rsons. The pressure o f event s was such that the full program must have been approved, fo r on 7 December 194 1, instead of hav ing onl y e ight y-five people, as wo uld have been the case had onl y the first incre ment been app roved. the strength of the Signal Inte lli gence Service was as fo llows:

35

THE ROA D TO PEA RL HARBOR

Officers Warrant Offi cers Enlisted Civ ilian Total

Washington 44

Fie ld I

Total 45

28 109 18 1

149

177 109 331

150

The staff was o rga nised as sho wn in the accompanying c ha rt , with the fo llowing sections and personnel: C ivilian 2 6

Total 4 8

10 I 16 4

24 15 10

34 16 26 4

5

2 15 22

16 22

109

153

Officers Administrative Personnel C ryptanal yti c Japanese German Cryptographic Laboratory Communications Tabulating ( IBM ) ·'Trainees" in classes Total

2 2

44

7

.. .The nineteen people in professio nal g rades a lso inc lude Japanese tra ns l,,tors; o nl y a few ex pe rt c ryptanal ys ts we re the refo re o n ha nd among those nineteen. The necessity for producing codes and ci phers has a lready been mentioned and the staff of 26 persons was hardl y adequate for the large volume of work to be done. The production of our e lectri c c ipher machines (S igaba) was unde r way and there were multitudino us detail s to be dealt with in that project above: the preparati o n of instructional manuals for the maintenance and operation of the mac hine, the e laboration of keying arrangements; the e laboration of what shall go into spare parts kits, etc . Also, the plans for the production of Convert er M-209 were unde r way- a very large project in itself. Besides a ll this, keys and control tapes for the e lectric Converter M 134A had to be prepared on a c urre nt basis, to mai ntain security of c urre nt communicati ons; strips an d key lists for the w idely distributed Strip C ipher Device M- 138A had to be prepared in large quantities; constant liai son with the Navy Cryptographic Division. for the prod uction a nd di stribution of cry ptog raphi c material for joi nt communicati o n required much time, etc . Also. plans for the production of an a ut o matic te le type enciphe rment system in vented by Signal Intel li gence Service personne l were being brought to fruition.

36

SIGNALS INTELLIGENCE

The ever-expand ing requirements of traffic for the cryptanal ytic staff meant expansion in the facilities and personnel of the Second Signal Service Company. Seven monitor stations were in operation by 7 December 1941 : I - Fort Hancock, New Jersey 2 - Fort Scott, Presidio of San Francisco 3 - F0I1 Sam Houston, Texas 4 - Post of Corozal. Panama Canal Department 5 - Fort Shafter, T. H. 6 - Fort McKinley, P. I. 7 - Fort Hunt. Virginia Also, of course, a good portion of our traffic came from the United States Navy, which had a large station at Bainbridge Island, Washington, and another at Bar Harbor, Maine. The mere recording and index ing of the large volume of messages

constituted a task of no small proportions. Hence, even with the large expansion of the Signa l Intelligence Service there was much more to do than hands to do wi th . Consider what was done in the field of solving and translating Japanese diplomatic messages alone. By a special agreement between the Chief Signal Officer and the Director of Nava l Communications, all work on Japanese diplomatic traffic was divided up between the Anny and Navy on the basis that Army would process all messages of even date (cryptologic), Navy, all of odd date - with full exchange of technical data and results. Thus, in collaboration with the Navy and with a total of only 34 persons, the Signa l Intelligence Service was attempting to decode, decipher. and translate its half of the traffic for which the Japanese must themselves have had several hundred persons, and the Japanese, of course. did not have to translate the material , a job which was quite difficult for us because of linguistic difficulties and paucity of qualified trans lators. Moreover, the Japanese had the necessary keys; we had to solve them , and in many cases so luti on had to await the accumulation of traffic - it could not always be done with a single message, so that occasion delays in solution were inevitable from the very nature of the systems involved. Also, another factor had to be taken into account: the time required to forward the intercept traffic to Washington. Lacking facilities and personnel to prepare all the traffic, encipher it, and transmit it to WaShington. some of it had to await shipment by air mail, and thi s factor alone occasioned delays. As has bcen aptly said by General Miles: "The astonishing thing, gentlemen, is not that these messages were delayed, but that they were ab le to do it at all. It was a marvelous piece of work." WILLIAM F. FRIEDMAN Director of Communications Research Prepared on 4 December 1945

THE ROAD TO PEARL HARBOR

37

From SRH 13 5 . Army success in sigllals illlelligellce aperatiolls prior to World War /I was depelldent 1I0t only all the core of skilled cryptanalysts at SIS headquarters, iJIII all the men who manned what we naw call the "frOllt end" of the system: the Signal Corps intercept operators. Th e Army's Chief Signal Officer, Major General Joseph O. MauiJorgne, describes the Army's existillg intercept organi:atiol1 alld suggests a way to improve it il1 this 1938 letter. Th e 2d Sigllal Service Company he proposed was activated 011 I January 193 9. WAR DEPARTM ENT Office of the Chief Signal Officer Washington OCSigO-320.3- RID

Se ptembe r 2, 1938

SU BJ ECT: Signal Intelligence TO: T he Adjutant Ge ne ral I. The Signal Intelligence Service in Washington depends for the material with which it works, upon Signal Inte lligence Detac hme nt s in Hawaii , Panama, Ninth Corps Area and Eighth Corps Area. These signal intelligence detachments are parts of the Si gnal Service Companies located in the same corps area or departme nt. There ha s been o rde red set up as pa rt of th e Te nth S ig nal Se rvice Company in the Philippine De partment , a signal intelligence detachment of 10 men. In addition , so important has the War De partment conside red the work o f the Signal Inte lli gence Service, it has been thought essential to detach 17 men from the Provisional Radio Intelligence Company at Fort Monmouth. New Jersey, and have this detachment ac t as a signal intelligence de tachment. This last is done even at the risk o f crippling the important tac tica l activities of the Prov isional Rad io Intelligence Compan y (to be designated as the I st Radio Intelligence Company as of November I, 1938). 2. Wh ereas th e res ult s ac hi eve d b y th e S ig nal Int e lli ge nce Se rv ice in WaS hington a re treated with utmost in care and secrecy, equal measures o f secrecy arc not taken with the personnel of the signal intelligence de tac hme nts in the fi e ld . T he personnel o f the signal inte lligence de tachments cannot long be unaware of the nature of the work in which they are engaged and the ultimate secrecy o f the syste m as a w ho le depe nd s ju st as mu ch upo n th e m as upo n the pe rso nn e l in Was hington. At present the cho ice of signal intelli gence detac hme nt personnel is left to the corps area and de partme nt signal o ffi cers and there is no doubt that they exercise every poss ible precaution. However. in th e present form of organi zati on

much is e ntirely o ut of the ir hands and beyond the co ntro l of the Chi ef Si g nal Offi cer. For exampl e, a skill ed membe r o f the signal inte ll igence de tac hment in Hawaii is re tumed to the mai nland . Unless he is in G rade 4 or above he is reduced

38

SIGNALS INTELLIGENCE

to the grade of private and returned for haphazard assignment by the normal process. Even if he is evelllually reass igned to a signal illlelligence detachment on the mainland , he must begin afresh in the grade of private and work himself up. Knowledge of thi s actuates many to separate from the service; every such man separated from the service canies with him a fairly accurate knowledge of the work upon which the Signal Intelligence Service is embarked and he constitutes just one more threat to the secrecy of the work . He is a far greater threat when separated from the service since he is no longer unde r military surveillance. 3. The haphazard movement of the enlisted personnel engaged in signal intelligence work also makes their training more difficult and lessens the effectiveness of the Signal Inte llige nce Service. The useful member of the s ig nal intelligence detachment must be first of all an excellent radio operator and beyond that must acquire sk.ill and experience peculiar ro signal intelligence work: the acquisition or this skill and ex perience requires from six month s to a yea r under competent instruction. At present, it is not possible to surely retain trained me n: indeed it is not even poss ible to surely retain compete nt instructing personnel. 4. The remedy to thi s situation is a signal service compan y, detachments of which are stationed in the various corps areas and departments as required. The personnel could then be transferred from detachment to detachment without loss of grade or rating and control over the personnel would be increased. To minimize the number or highly skilled personnel separating from the service, the grades and ratings of the compan y should be considerably higher on the average than is nornlal in the Army. After many years of experience, the Navy Department adopted just the form of organization here recommended. Obviou sly, the idea l ca nnot be achieved by me re statement. However, a great step forward in the secrecy of signal intelligence work and in the training or signal illlelligence e nlisted personnel can be taken by organizing the several signal intelligence detachments into a signal service company. This can be done without request for any additional men, grades, rating, housing, transportation or equipment.

5. It is therefore recommended that there be organized the 2d Signal Service Company. The personnel necessary are to be obtained by withdrawing from the various servi ce companies the signal intelligence detachments of which are now

part of them and by a temporary decrease in the I st Radio Illlelligence Company. ( It is proposed not to alter the Tables of Organization of the Radio Intelli gence

Company, but mere ly to leave vacant such grades and ratings as are taken to make up the 2nd Signal Service Company: when increases in signal intelligence personnel are granted the Radio [Illelligence Company will be permitted to rega in its table of organization stre ngth.) Inclosure No. I, Table of Organization of the 2nd Signal Service Company, shows the strength , grades and ratings of the proposed company. Inclosure No.2 shows the men, grades and ratings taken from the various companies in order to constitute the Signal Intelligence Company. It will be seen that the organization of the Signal Intelligence Company requires no addition-

39

THE ROAD TO PEARL HARBOR

al men, grades o r ratings from outside the Signal Corps itself. The matter of housing wi ll not be concerned since the detachments are already housed. No additional transportati o n or equipme nt w ill be required because of the organi zati o n of the compan y. The company w ill be comm anded by the officer in charge of the Signal Intelli ge nce Detachment at Fort Monmo uth . Detachments of the compan y will be stationed in the vario us corps areas and departments just as they are now : they wi ll be attached to the local organi zatio n for rations and quarters (t he same o rganizations w ith which they a re now rationed and quartered): the ir work wi ll be supervised by the corps area o r department sig na l officer concern ed, except in the case of the detachment now at Fort Monmouth which wi ll be supervised by the com manding officer of the com pany. 6. It is intended that the 2d Signal Service Company shall function unde r general ins tru cti ons issued by the Chief Signal Officer. Plans for its expansion in time of war and its use in wa r will be recomme nded by the Chief Signal Officer fo r inclusion in the severa l co lored war plans, and the mobilization plan. /s/ J .O. MA UBO RGNE /t/ J.O. MAUBORGNE Majo r General C hief Signa I Officer of the Army

From The "Magic" Background of Pearl Harbor, Volume IV, Appendix (Was hin gton, D.C. : De partment of Defen se, 1977), p. A-89.

The Japanese government di~patched this diplomatic communication to its envoys in Washington, Special Envoy Saburo Kurusu and Ambassador Admiral Kichisaburo Nomura , 0 11 22 November 1941 . It was illlercepted and translated by the Army's Signal illtelligence Service 011 the same day. In the aftermath of Pearl Harbor, its lallgllage took 011 a lIew omillOIlSlless. FROM: Tokyo TO: WaShington November 22,1941 (Urgent) #812 To both your Ambassadors. It is awfully hard for us to consider Changing the date we set in m y #736a. You should know this. however, I know yo u are working hard. Stick to yo ur fixed policy and do your very best. Spare no efforts and try to bring about the solution we desire. There are reasons beyond yo ur ability to guess why we wanted to sett le Japanese-American relations by the 25 th . but if w ithin the next three or fo ur days

SIGNALS INTELLIGENCE

40

you can finish your conversations with the Americans; if the signing can be completed by the 29th (let me write it out for you-twenty-ninth); if the pertinent notes can be exchanged. if we can get an understanding with Great Britain and the Netherlands; and in short if everything can be finished, we have decided to wait until that date. This time we mean it, that the deadline absolutely cannot be changed. After that things are automatically going to happen. Please take this into your careful consideration and work harder than you ever have before. This, for the present, is for the infomlation of you two Ambassadors alone.

***

Trans. 11/22/41 (S)

ARMY671025138

From SRH 115. This excerpt from hearings held by the Army on the roots of the intelligence failllre that led to Pearl Harbor demonstrates all too vividly the feeble resources of the Military Intelligence Division on the eve of war. It also shows how tightly "Magic" was held: Brigadier General Hayes W. Krone!', the head of MID's Intelligence Branch in 1941, was unaware of its existence. This is the transcript of an interview of Kroner conducted in September 1944 by Colonel Carter W. Clarke, Deputy Chief of the Military Intelligence Service. The General Miles mentioned in the text was the hapless Brigadier General Sherman A. Miles, Assistant Chief of Staff G-2, at the time of the Pearl Harbor allack. "Minxlel;" whom Kroner describes as the individual who fi~l'I1ished mysterious source materials to one of Kroner's subordinates, was actually Colonel Rex A. Minckler, Chief of the Signal Intelligence Service at the time. Place: Date:

Time:

Room 2C637 Pentagon Building, Washington, D.C. 13 September 1944 1430- 1510

Present: Brigadier General Hayes A. Kroner Colonel Carter W. Clarke LI. Col. E.W. Gibson Brigadier General Hayes A. Kroner, having been sworn and warned of his rights by Colonel Carter W. Clarke, gave the following sworn testimony: Col. Clarke: Will you state your name, rank and station. Gen. Kroner: Hayes A. Kroner. Brigadier General. Station Rio de Janeiro,

THE ROAD TO PEARL HARBOR

Col. Clarke : Ge n. Krone r: Co l. Clarke: Ge n. Kroner:

Col. Clarke: Gen. Krone r: Col. Clarke : Gen. Krone r:

Col. Cl arke: Ge n. Kro ner:

41

Military Attache. Whe n did you first come on duty in G-2, De pa rtment General Staff? About the first of March 194 1. What we re your duties at that time? I was assigned to the Briti sh Empire Section of the Intelligence Branch, G-2 but before taking up those duties I was sent to Eng land for a pe riod of observation o f the war. I re turned from England earl y in June, a bout the middle of June 1941 . and took up my duties of Chief of the British Em pire Section, G-2 at that time. How long did yo u continue on that dut y? Approx imately one month. And then what di d yo u do? About the middl e o f Jul y I was infornled by the re tiring Chief of the Inte lli gence Branch, Co l. C. H. Mason, that he was leaving G2 and that he had been instructed to turn the Branch over to me as next senior officer. I immediately took over charge of the Branch and for the nex t several months was ac ting in charge and late r on I was appointed Chi ef of the Branc h. To the best of my me mory it was about Se pte mbe r 17 that an order was issued by the Executive Orficer, G-2 appo inting me as the Chief of the Intelligence Branc h, whi ch position I held until some time in the month of Decembe r. whe n Gene ral Lee became Acting Assistant Chief of Staff, G-2 and I was then verball y relieved from Chief o f Intelligence Branch a nd made Deputy to Ge neral Lee. What did the Intelli gence Branch consist of principall y? I found the Intelligence Branc h, in Jul y ' 41 , to consist of a ve ry small Admini strati ve Section and of seve ral geographic sections a nd an Air Section . The Admini strative Section was in process of transition . Becau se of the impac t of the wa r and more informati on

coming into G-2, the re was a greater need fo r prope r handling and di sseminati on of that information. The onl y geographi c secti on of G-2 whi ch appeared to be ex panding at that time was the Latin American Scction. By the 10th of Octobe r- refe rring here to the offi cial c ha rt on record- the Intelligence Branc h consisted of a headquarters or the Admini strati ve Secti on, Contac t Secti on. Situati on Sec ti on, Di sseminati on Sec ti on. and the Air and geo-

graphic secti ons composed of the fo llow ing: Air Section. Briti sh Empire Section. Western European Section, Central Euro pean Secti on. Eastern European Section, Southe rn European Secti on. Far East Section, and the Latin A me ri can Secti on. T he several

42

SIGNALS INTELLIGENCE

European sections had been set up to deal with the increased inFo rmation coming in from the European war and, together w ith th e

British Empire Section, it was conside red adequate to handle the war in Europe. The Latin American Section was ex panding fo r o bvio us reasons o f hemi sphe re de Fe nse. The Far Easte rn Section at thi s time had not unde rgone any mate ri al change, so far as I can remember, for several months. Thi s organi zati on, w ith a few

Col. Clarke :

Gen. Krone r: Col. Clarke : Ge n. Kroner: Co l. Clarke :

c hanges, continued during my tenure of ofFice as Chief o f the Branch... . I unde rstand you to say the n that yo u were Chief of the Intelligence Group from about Jul y through Pearl Harbor, Decembe r 7? Yes. During Ihal time was the re an y recommendation made- first let me ask you thi s-who was the Chief o f the Far Easte rn Section? Col. R. S. Bratton . During that tim e was there an y recomm endation made to you as

Chief of the Group fo r an y increase in pe rsonnel or ex pansion of its intelli gence gathe ring ac ti vities? Gen. Kroner:

Col. Cl arke : Gen. Krone r:

Col. Cl ar ke:

Ge n. Krone r:

N ot an y to give me an y trouble - th ere was some minor increase of

a cle rk or two and perhaps one officer, I don ' t re membe r, but nothing in the line o f ex pansion as I have re fe rred to in the European Section. I wo uld say this furthe r. that one of their Far East ex pert s.Col. Petti grew, was taken from the Far Eas te rn Section at the time to ass ist in the reorgani zation of the Headquarters di ssemination. so that I had very close to me at that time an ex pe rienced Far East officer on my staff. Who was that? Pe tti grew. It was, I think , worrhy of note just he re that Co l. Be tts, who was the Chief o f the Situation subsecti on. whom I call ed my G-2-G -3 offi cer on my own staff, Col. Pettigrew being so-ca lled G- I-G-4 offi cer, together with my own ex perience in the Far East. th at w ith Col. BraLton it was a rairl y good team of Far East trai ned offi cers at the head o f the Intelli gence Grou p. Was the re any ac tion ta ken by the Chief of the Far Easte rn Gro up that wo uld lead you to think that any diffic ult ies or hosti lit ies were ex pected in the O rient that wo uld affect the United States? None spec ifica ll y as to direct threat in regard to time or date or pl ace. Japanese possible lines of ac ti on were oft e n di scussed , as yo u mi ght expect amo ng officers as closely allied as the ones lmenlioned above w ith Far East training and influe nce. I left the evaluation of inform ati on pertai ning to the Far

43

TH E ROAD TO PEARL HARBOR

Col. Clarke: Gen. Krone r: Col. Clarke: Gen. Kro ner: Col. C larke: Gen. Kro ner:

East and it s interpretati on initially to Col. Brallon. a nd in that over-a ll sense to Co l. Bells. What was your chief source of in fo l111ation w ith regard to the Far East? Mil itary Allache reports and translations of books and journals which were received from vari ous sources. so far as I was aware. You had no secret intell igence service? No. Wa s any atlempt made to create any during the time that YOll we re in c harge? Yes. I remembe r the time- I do n' t remember exactl y- but I re me mber an instance whi ch was pal1ia ll y ful filled , I be lieve. that a fo rme r Far East tra ined officer. who I th ink was retired and li ving in California, Capt ain or Majo r W. C. Clear, was brought into G-2 and give n sec ret brie fin g by Gen. Miles. to the best of my knowledge. and sent out to the Fa r East. I think to Si ngapore, in order to initi ate some secret inte lli gence, of which I have no detai led know ledge.

*** Col. Clarke: Gen. Krone r: Col. C larke: Gen. Kro ne r: Col. C larke: Gen. Krone r:

Col. Clarke : Gen. Kroner:

Did yo u have access 10 a source of informati on whic h we know as Top Sec re t or the British known as Most Sec ret? Meaning communications infol111ati on? Signal intel li gence. No. none whatever. You mean you d idn 't get it or yo ur Branc h di dn 't get it? I personall y as C hie f o f the Branch did not get it. I was awa re that something, whic h later I fo und out 10 be of thi s nature. ex isted. but I was g ive n to unde rsta nd. particula rl y by Col. Brall on and Col. Pe lli grew. who some times ha nd led the mailer fo r Co l. Brallon. that he received informati on from Co l. Minxle r. whom I knew to be in the S ignal Corps. which pe rhaps had to do with Japanese tfOO P movement s, whi ch he by long custom and by Ge ne ral Miles spec ia l des ire, was to handle himself directl y with Ge n. Miles. The n you do n' t know what Col. Brallon d id with th is stuff other than g ive it to Gen. Mil es? No offi c ial infol111ation. I freque ntl y remembe r seeing him leave hi s offi ce w ith severa l parcels unde r hi s arnl and be go ne fo r some hours, but I fe lt it was Illy dut y to fo llow the establi shed procedure whi ch was appare ntl y pleasing to Illy c hie f, and I did n't questi on

44

Col. Clarke:

Gen. Kroner:

Col. Clarke:

Gen. Krone r: Col. Clarke: G en. Kroner:

SIGNALS INTELLIGENCE

the procedure . During the time that you were Chief of the Intelligence Group you never had access or never saw an y of thi s mate rial yo urself? Whc n did yo u first scc it ? I only saw the mate rial , which later I learned was called material , when I used to receive it from Minkle r 's hands when Bratton was absent, and lock it up in my sa fe and give it to Bratton without sorting it out. In other words, I unde rstood, without any specific orders from Gen. Miles, that he wished it handled that way and the refore I did not violate that procedure. Do yo u know if any estimates were written in the Far East Scction based on this material or an y prognosti cations, forecasts, or predic tions? I do not. If any had been made yo u would ha ve seen the m, wo uldn't yo u, as the Chief? I think so becau se estimates were accustomed to be ing made, not

onl y at regular intervals. but also spec ial es timat es from lime to

time would be made, the Far Eastern part of which was always prepared initially in the Far East, held by Col. Bratton or under hi s direction brought i11l0 Col. Belts, who rev ised it and fitt ed it in with information from other geographi c sections. I assumed that the Chief of the Far Eastern Section used all the information at hi s disposal to make a complete estimate and as accurate an es timat e

Col. Clarke:

as poss ible. Did there exist anyt hing like a central evaluating section where all information came in and was th ere melted into one es timat e or report or summary?

Gen. Kroner:

No. There was a trend toward that at the end of the year but we didn't get a nyw he re with it. It was just in the planning stage.

:f:** Col. Clarke:

I would like to ask you one more question. In any estimate from the time yo u took over the Inte lli gence Group up to and including Pearl Harbor, was there ever an y prediction or forecast made of a poss ibl e attack on Pearl Harbor? Gen. Kroner: None to my knowledge. I have in mind the last estimate that was made before Pearl Harbor, which was a n estimate covering a future period from December I to sometime in 1942. Lt. Col. Gibson: Did you consider it a capabilit y of the Japanese to successfull y attack Pearl Ha rbor with bombe rs?

TH E ROAD TO PEARL HARBOR

Gen. Kroner: Co l. Clarke: Ge n. Krone r:

45

No. Thc mane r was di sc ussed Did you identify thi s document ? Yes. I identified it - thi s is thc docume nt to which I re ferred - I B 159, Novcmbe r 29, 194 1. This panicul ar estimate was conside red by the who le di vision. not onl y the Inte ll igence Group but by Ge ne ral Mil es himselr. as pe rhaps the most important we had ever gonc n out. That imponance lay not in so much the danger that we saw from Japan , although dan ge r in that fi e ld was pre tt y thorou gh Iy di scussed, but primaril y bccause Gen. Miles wished to focus War Depanme nt thought on the de fe at that could be admini ste rcd to thc Nazi powers. In the prepara ti on o f the estimate each geogra phic section in the Inte lligence Branch pre pared its part. Colone l - now Brigadier General - Thomas J. Ben s put the several estimates togethe r and d id what we ca lled " po li sh the m up."' He and I di scussed the lines of ac tion a nd the capa bilities o f all the warring powers and especially of each potenti a l e nemy to the U.S. A .. and I took the m to Ge n. Miles whe re they we re finall y alte red to suit him or approved. Thi s panic ul ar estimate does not incl ude in the lines of ac tion open to Japa n, a n attac k on Pearl Ha rbor, and I re member that so distinc tl y because when the word came through the radi o on that fat e ful Sunday, December 7, that Japan had attac ked Pearl Harbo r, I was sitt ing in my o ffice in the Munitions Building reading from thi s paper the Japanese capabilities. The rcfore from my po int of vie w, I feel that Japan 's potential capabilit y agai nst Pearl Harbor was left from thi s estimate because

Col. C larke:

Gc n. Krone r: Col. C larke :

ne ithe r Col. Be tts nor I had any infonnation whic h would lead us to be lieve that they we re capable o f or planned to do so. I would like to ask one fin a l question again just to re iterate the fac t that yo u personall y had no know ledge of what Co l. Bran on did wi th thi s most sec ret materia l or to whom he showed it ? That is correct, except to Ge n. Miles. And yo u don ' t kn ow what Miles did w ith it.

CHAPTER III Accepting the Challenge The technical success of American cryptanalysts could not prevent the Pearl Harbor debacle: the intercepted Japanese diplomatic communications had not contained the details of Japanese military planning. In the aftermath of Pearl Harbor, American policymakers realized that some better way had to be found to produce finished intelligence from COMINT intercepts. Secretary of War Henry L. Stimson called upon the services of the distinguished Chicago lawyer Alfred W. McCormack to solve the problem. McCormack recommended the establishment of a Special Branch within the Military Intelligence Service to handle highly sensitive signals intelligence. This proposal was adopted, and McCormack was commissioned as a colonel and became deputy head of the new organization. McCormack tells the story of the inception of the Special Branch in SRH 116, Origin, Functions and Problems of the Special Branch. The excerpt from SRH 141, Papers from the Personal Files of Alfred McCormack, Colonel AUS, Special Branch , G-2, Military Intelligence Division, War Department, Volume 1I , details some of the problems surrounding the effort.

48

SIGNALS INTELLIGENCE

From SRH 116. Writing in 1943 to his superior, Colone! Carter W. Clarke,

the Chief a/Special Branch, Colonel Alfred McCormack describes the orgal1i:ation they both had helped to shape.

OR IGIN, FUNCTIONS AND PROBLEMS OF THE SPECIAL BRANCH . M.I.S . April 15. 1943. MEMORANDUM FOR COLONEL CA RTER W. CLARKE: Subject: Origin, Functions and Problems of the Specia l Branc h, M.I.S . Following our recent discussions, I am pUlling down in this memorandum what seem to be the important facts relating to the Special Branc h. in the hope of clarifying the problems now confronting us. A. Respo nsibilit ies of Wor Department for Int elli gence Derived from

Cryptanalysis. The outstanding feature of "black c hamber" activities in the United States is that they are performed almost excl usively by the anned services and that the great bu lk of the wo rk is done by the Army. The FB.I. and the Coast Guard operate in certain limited fi e lds. The Navy confines itself excl usive ly to na val traffic. The Army is responsi ble not only for military and MIA traffic bu t for the large residual fi eld that is loosely ca lled "d iplomatic," which includes the great bulk of the impo rtant traffic over comme rcial , official and semi -offic ial radio networks throughout the world. [text withheldl From the standpoint of the Army the present U.S . arrangements a re the most desirable, especially in time of war. The Army is not depe ndent upon the State Departme nt, [text withheldl for its information about all the complicated cross-currents of political and economic diplomacy and for the enormous amount o f intelligence that is found in foreign officia l messages. On the contrary, the State Department is dependent upon the Army. which now contro ls the so urces a nd determines the order of priority in which those sources will be exploited. The continuation of Army-Navy control over such sources is important for the immediate necessities of war, and

it is al least eq ually important for the long-range

future. The availab lilit y of important intelli gence from cryptanalytic sources. during the c ritica l 18 months prior to December 7, 1941 , was due solely to the fac t that the Anny and the Navy, in the face of difficulties and even legal prohibitions. we re able to maintain ske le ton organizations in thi s field , in preparation for the day when they wou ld be vita l to the national defense. The armed services are the safest repositories of cryptanalytic material and act ivities, since they can operate with greater

ACCEPTING THE CHALLENGE

49

secrecy and with a longer-range viewpoint than can the politica l agencies of the Government. The chances of cont inued cryptanalytic progress after the war are believed to be better if control of the fi e ld remains in the armed serv ices than if the non-military fi elds are taken over by a civilian agency. However, control invo lves responsibilities, and in case of the diplomatic traffic it involves specific commitments to the President and to the State Department to get as much usable intelligence as possible out of the material. B. Handlin g oj 'I7Tercepr 'I7Telligence Prior ro rhe War. When this country declared war, both the intercept work and the cryptanalyti c work in the diplomatic fie ld were divided between the Army and the Navy, though actually most of the work was being done by the Army. The combined intercept and cryptanal ytic facilities of the Army and Navy had produced some 7,000 decrypted and translated messages in the 6 months before Pearl Harbor, or an average of about 300 per week. of which onl y 1.561 were considered of enough signifi cance to be fil ed for future reference . The responsibility for bringing important messages to the attention of the Secretary of War and the top officers of the Army was vested in the Chief of the Far Eastern Section , M.!'S. O.N. !. served the President , the Secretary of the Navy and the top officers of the Navy. In M.I.S . the material was handled as follows: Each day's batch of messages was exam ined by the Chie f of the Far Eastern Section . Those not considered important were put aside for burning. Those that were thought important were c irculated to the Secretary of War, the Chi ef of Staff, the Assistant Chief of Staff, Op.D .. and the Assistant Chief of Staff, G-2. One copy of each circulated item was put into a locked file; the other copies were bumed after they had been retumed by the readers. No personnel were ass igned to a continuous study of the material ; hence very little could be done to put any of it together in connected form . The Chief of the Far Eastern Section, M.I.S., having as such a full -time job and limited personnel, could do no more with the messages than carry the important facts in his head. C. Decision oj rhe Secrerary oj War Concerning Exploirarion oj Crypranalyric Field. When the sudden attack on Pearl Harbor occurred, it became apparent that the event had been clearly foreshadowed in the Japanese traffic of 194 1. The Secretary of War. and no doubt others. than concluded that the traffic had not been g iven suffic ientl y close attenti on, and that some agency should be set up to deal with cryptanalyt ic intelligence in a more thorough-going fashion than previously had been thought necessary. The Secretary decided that the job ought to be undertaken by a lawyer hav ing a spec ial type of competence and training, such as may be acquired in the handling and presentation of large cases in volving complicated facts. After consulting with the Assistant Secretary, Mr. McCloy, about possible candidates, the Secretary authorized Mr. McCloy to offer thi s assignment to me. my experience over a number of years hav ing been along the indicated lines. The mat-

50

SIGNALS INTELLIGENCE

ter was put up to me early in January, 1942, and after some discussion I agreed to undertake the job, resigned from my law firm, and took office as a Special Assistant to the Secretary on January 19, 1942. On that day I received my instructions from Mr. Stimson. Briefly, they were that I was to study the problem and determine what had to be done in order to expand the cryptanalytic operations to meet the requirements of war, and to make sure that all possible useful intelligence was derived from the cryptanalytic material. Both the Secretary and Mr. McCloy emphasized the need for close and continuous study of intelligence derived from cryptanalytic activities and of piecing such intelligence together in connected form. For two months I studied the back materials, investigated production facilities, had numerous conferences with Mr. McCloy, General Lee and others interested, and made recommendations as to what should be done and how the work should be handled. The details are no longer of any importance. Everyone agreed that there was a very large job to be done. D. Olganization of M.l.S. Section to Handle Intercepts. Shortly after I started work, there was set up in the Far Eastern Branch of M.I.S. a section which was to follow and report upon the cryptanalytic material. It soon became apparent that there were at least two fundamental differences of opinion between the officer chosen to head that section and myself. First. I disagreed with the notion that any reserve officer, or any civilian who

had been graduated from college, was qualified to handle cryptanalytic intelligence. It seemed to me that the job could be done effectively only by persons with very special qualifications; that the most careful selection of personnel was necessary, and that only individuals of first-rate ability and suitable training should be taken into the work. Second, I disagreed with the method that was adopted for reporting the intercept material, which was to take what looked interesting and to pass it along in paraphrased form, without any attempt either to check or evaluate the infornlation or to supplement it by collateral intelligence. It seemed to me that the Secretary, the Chief of Staff and the others to whom reports were made were entitled to have every item carefully checked, evaluated and supplemented by all possible sources of intelligence, and that their time should not be wasted in reading odd and unchecked bits of information not related to attendant circumstances and given their proper value. Further, it appeared to me that the daily reporting of cun'ent messages was only one part of the job; and that the real job was to dig into the material, study it in the light of outside information, follow up leads that it gave, and bring out of it the intelligence that did not appear on the surface. E. Organi:arion of Special Branch. My investigations of the intercept and cryptanalytic problems brought me in contact with you, and we had many discussions on the subject and some discussions with General Strong. His thinking was along the same lines as ours and he

ACCEPTING THE CHALLENGE

51

approved of our conclusions. The upshot was that the M.I.S. section in charge of cryptanalytic activities was made the Special Branch , with you as its Chief. We moved into common quarters and worked together, your staff (part of M.I.S.) and my staff (part of the Secretary's office) being made into a single working unit as fully as if they were all part of the same agency. You concent rated on pushing the ex pansion of the intercept faciliti es, on setting up a workable system of priorities, and on the man y technical problems that were encountered. You were full y famil iar with those problems and had arri ved at the same conclusions that I was reaching, with respect to the necessity for a great expansion of the intercept faciliti es and for an attack on the cryptanalytic problems on a much broader basis. You were largely instrumental in inducing the Signal Corps to embark upon the program that it is now carrying to completion , invol vi ng the acquisition and development of Arlington Hall . the locat ion at Vint Hill of one of the largest intercept stations in the world. the expansion of facilities at Two Rocks, Califomia, and elsewhere. and a personnel program reasonably adapted to requirements of a war-time job. Meanwhile, I concentrated on recruiting a suitable staff and on the problem s involved in getting int e lligence out of the intercepts and devising methods of check ing, evaluating and reporting it and of keeping track of it by suitable files and indexes. General Strong and yo u felt that it was desirable for me to go into the Army, and I was commiss ioned in June. Thereafter the Special Branch absorbed the staff that the Secretary's office had assigned to me. F. Work oflhe Special Brallch.

*** It seems fair to say that we have a good organization of people well above the average level of competence, who work full steam and on the whole have given a good account of the mse lves. Our so-ca ll ed " Magi c Summary" appears to be accepted as a satisfactory job of reporting. At the Navy's request a copy now goes to the Secretary of the Navy, and at the State Department 's request copies no w go to the Secretary of State and to the Assistant Secretary of State who follow s cryptanalytic intelli gence. A great deal of our work does not get into the " Magic Summary." I think we probably do more wo rk in deciding what to exclude from the Summary than we do on items that are put into it. We try to avoid reporting more details. and we consider that those to whom we report should not be bothered by material which, if properl y checked and studied, wo uld be found to be of no value. A lot of our work, from the reporting angle, consists in feeding information to the M.I.S. geographical sections, to other agencies of the Department, and to outside agencies. We employ one officer full -time in verifying Order of Battle information and reporting such information to the O/B sections and the Air Branch of M.I.S. When securi ty considerations pemlit. we pass on economic information to

52

SIGNALS INTELLIGENCE

the B.E. W. We are constantl y called upon fo r infonnati on by those in the State De panment who have access to our material, and we employ one o ffi cer full time in serving the State Depanment and gelling necessary information from them. We prepare periodica l re ports on South Ame rican information fo r the Coordinator of Inter-A me ri can Affairs. Whe n information is o f a sort that the Ass istant Chief of Staff, G-2, should bring to the atte nti on o f othe r agencies. such as the O .S.S. o r the O.W.I. , we pre pare reports or pa raphrases for hi s signature. The who le body of intelligence whic h nows out o f inte rcept material, int o the various sources that have been mentioned, adds up to a considerable amount. Thi s memorandum will not allempt to evaluate it , as compared with other sources o f intelligence, but it may be wonh while to me nti on a numbe r of examples of importa nt intelligence material , with whic h the reade rs of the " Magic Summary" will be famili ar: a. Practicall y the onl y good informat ion abo ut the relations bet ween Japan and Russ ia has come from the intercept mate ri al. Much of it has not been obvious on the face of the mate rial , but has required a g reat deal of stud y and collateral research in o rder to piece it together. (A t a time whe n almost everyone expected Japan to attack Russ ia, and when M.I. S. we nt o n record as predicting such an allack at an earl y date, we had concluded from the inte rcept materi al that Japan had no intenti on of allacking Russia, but wanted to keep out o f wa r wi th Russ ia at all costs.) b. The best in fo rmation about the allitude a nd inte nti ons [text withheld ] follow ing the African in vasion, came from that source. c. That source has also yielded. in addition to conside rable infollllati on not othe rwise obtaina ble abo ut Japan. the most acc urate and de pendable info rmation about the relatio ns between the Japanese Gove rnme nt and its Axis partne rs and abou t trade between Axis Europe and the Far East. [text withheld] Many other examples of lesser impo nance, but of the same general type, might be given. It wo ul d take a lot of space to g ive a good picture of what might be called the "detail inte lligence" that we get, of its present va lue and of the future va lue that it may have as it graduall y builds up fac ts. Perha ps the picture can best be g iven by the fo llow ing illustrations: T he nat ure of the materials bearing on the Far East is such that the most as tute pe rson, even if an ex pe n on the Fa r East and possessed o f a photograp hic me mo ry. wo ul d de ri ve few connected impressions- and ve ry lilli e usable in fo rmati onfrom merely reading the messages from day to day. T hey m ust be pieced together by the most painstaking and labori ous process, involvi ng coll ate ral in vesti gati on. ofte n o f minute po ints. The work is ve ry diffic ult , but the rewards in the way o f intelligence are substanti al. For example : As you know, we are in possess ion today of enough informati on abollt

Japanese shi ppi ng to and from Indo-China and Thail and so that we have a prelly

ACCEPTING THE C HALLENG E

53

clear picture of the movement of supplies on those runs and of the proble ms that Japan is encounterin g in getting raw material s out of th ose areas and keeping them

supplied with the minimum of essenti al import s. Our info rmation has been put togethe r from an infinite number of scraps o f fact. As yo u recall , the mate ri al on thi s subject was not at first very promi sing. It looked hope less to make any connected sense out of the garbled texts. incomprehe nsible ship names, abbrev iations and alphabetical des ignations, and the inconsiste nt tonnage fi gures, loading re ports and ship schedules, especiall y beca use in many cases the key messages had not been intercepted. Finally, howeve r. by struggling with the S.S .D. translators. doing research on Japanese merchant vesse ls. graduall y ide ntifying ship names, eliminating duplicati ons o f cargo information. checking ship tonnages and schedules and , in gene ral, appl ying what might be called high class detecti ve methods. we built up a picture that is clear and substantially complete in its principal parts. Hav ing built up that picture, it is now poss ible for us to fo llow the pe rtinent messages from week 10 week and to unde rstand what they mean. Had the job not been done. the c urrent messages wou ld be largely incompre he nsible. Since Japanese shipp ing has been selected as a n illustrati on o f the pi ec ing together of incompre hensible bits of intelligence into a connected who le. it is inte resting to note that we arri ved at several conclusions about Ja panese Shipping which made

it necessary to change prev iously accept ed fi gures. For instance. we estab-

lished that the Japanese ha ve bee n employing at least 450,000 tons o f merchant Shipping in the South Pacifi c, whe reas the J.I.c.. apparentl y with the approva l o f the Na vy, recentl y publi shed a study which showed that onl y 300,000 tons of shipping we re e mployed in moving goods to and from that area. We brought our find ings to the attenti on o f O.N.I., whose peopl e first expl ained them away. A few days ago. however. we we re ad vised informall y by O.N. 1. that they no w agree with our findin gs and that, while they do not want 10 rev ise their publi shed fi gures at present . they will make the correcti ons whe n they nex t issue a report. T his little slOry has a pe rtine nt moral. Until we gave them the above and certain othe r find ings on Japanese me rchant shipping, the O .N. 1. people were polite and cooperati ve, but they SlOpped short o f givi ng us their own secret information. But , whe n we bro ught the m to our office and ex plained what we we re trying to do and how fa r we had prog ressed. they o ffe red to put at o ur di sposal all their own infomlal ion on the same subject. Since that time, which was last fall , they have pre pared for o ur special use and delive red 10 us each morning a summary of all information received on the prev ious day about me rchant shipping in the Far East, fr0111 sources other than our own material.

From th is dail y re port . pieced together with our other mate rial , we are building up a wealth o f ev idence on the moveme nts of Japanese vessels and cargoes. We have hopes that the time is not far di stant when we will ha ve rhe whole pi cture o f Japanese shipping and wa ter-borne trade reconstructed to a po int whe re irs main

54

SIGNALS INTELLIGENCE

outlines will be clear and many if not most of the important details will be known. The above is given as a sample of the work that we are trying to do; and while this memorandum is not intended as an analysis of S.S.D. traffic, present or potential , it may be worth while to carry the story just a little further: Japanese shipping is only one of the important subjects on which the intercepts are gradually building up connected and definitive information, not obtainable from any other source. Referring only to the Far East, it is our hope that over the next year, if the same kind of material keeps coming in and if sufficiently painstaking work is done on it, we will have the Far Eastern picture---except perhaps for the East Indies-pretty well reconstructed in its economic and political aspects and be able to make increasingly accurate diagnoses of Japanese capabilities, difficulties and plans. The story of the Dutch East Indies is buried, we think , in the water transportation traffic of the Japanese Army. The cryptanalysts at Arlington Hall think they have solved the cryptographic system in which this traffic is sent and that in 1,2 or 3 months they will have built up code values to a point where the traffic will be completely readable. If that should prove to be true, then the East Indies will be opened up from the intelligence standpoint and the geographical coverage of the Far East will be complete. G. Increase in Volume ofS.S.D. Material. Meanwhile. coverage is being extended, new systems are being solved and translation and processing facilities are being built up. [text withheld] In the 6 months before Pearl Harbor almost all the traffic produced was Japanese, with a scattering of Gernlan [text withheld] messages. A large volume from many sources is now handled, and language coverage has increased [text withheld] The recent solutions of the German Key Word Code and the Japanese MIA Code represented substantial progress in cryptanalytic work; and the expected solution of at least one of the highest grade Japanese military ciphers-and possibly of the whole Japanese military system-will, if it occurs, be an event of major importance from the intelligence angle. Intercept coverage has been extended by the development of 3 large fixed monitoring stations, giving approximately global coverage for what is broadly described as "diplomatic" traffic, and by field radio intelligence companies. Not only do the new stations permit more intercept facilities to be used, but the better antenna installations and better equipment make it possible to listen to circuits at times when previously they could not be heard. For example, until completion of a recent installation at Vint Hill , the Berlin [text withheld] circuit could be listened to for an average of perhaps one or two hours per day. Now it is estimated that the good listening hours have increased to 8 to 10 hours per day, and possibly more. When the Fairbanks station-the only one that can listen to traffic between Tokyo and [text withheld] -gets new antennas, as it will when the spring weather comes, the amount of traffic intercepted from those circuits should increase frolll a mere

ACCEPTING THE C HALLENGE

55

driblet to a substanti al volume. During the mo nth of March we received 4,500 dec iphered me ssages from Arlington Hall. This was still only a fraction of the available mate rial , as may be shown very readily by the figures for the messages [text withheld] during March.

Japanese AmlY and weather traffic Other traffic

Total 9 1,596 22.489 114,085

The bottleneck-by reason of which thi s Branch received onl y 4,500 messages from Arlington Hall in March, as compared with our 22,000 (excl uding Japanese military) [text w ithh e ld] li es in th e processi ng and tra ns latin g facilities of Arlington Hall , which are largely a problem of civilian personnel. They had 105 civi lians on Decembe r 7. 1941; they had 1,754 on February I, 1943 (the last date whe n we obtained exact figures). They now have over 2,300 and the ir TIO calls for 3,683 by June 30, 1943. It is estimated that- apart from the problem of Japanese military tarffic- the volume of S.S.D. material coming into the Special Branch wi ll at least double by the end of June. It will continue to increase as the successive bottlenecks are el iminated. More traffic will be inte rcepted; more will be decoded; more will be translated. Leav ing out o f account the Japanese Anny traffic, the total volume could easily increase four-fold by the end of the year. This inc rease of material e mphasizes our manpower problem. But the problem exists independently o f the question of volume.and the factors involved in longterm pl anning are much more important than merel y the question of handling a certain number of intelligence items per day. H. Basic Factors in the Problem ojCry{Jtanalytic Intelligence. Certai n cons iderati ons that are fundamental to o ur problems may be mentioned: a. An effecti ve intelligence age ncy--espec iall y one that must deal with such difficult material as decrypted intercepts-cannot be built up by writing a TIO and filling the places with what personnel happe ns to be avai lable. To do the work well , a man must have not only a broad educat ion and background of infonnation , but must have more than his share of astuteness, skepticism and desire to solvc puzzling problems; and he must have a capacity for laborious detail work that very few people have. As you know, we have had 2 cases of men who proved incapable of doing our wo rk, though both of them had had fairly successfu l careers in civil life. Not on ly were they able to produce nothing by their own efforts. but they used up so much time of others worki ng in the office, and made so many mistakes that it took time to correcl. that we fo und our net efficiency increased by taking them entire ly off the work and in vent ing jobs to keep them busy until they could be assigned elsewhere.

SIGNALS INTELLIGENCE

56

b. One principal reason why exceptionall y qualified personnel are required in the intelli gence agency that deals with cryptanal ytic material is that such an agency cannot afford to make mistakes, because it tends to become a focal point in intelli gence activities. The re a re several reasons for thi s: ( I) Intelligence from inte rcepts, where the material is first-hand and of de pendable origin, is the most reliable infoll11 ation and frequently the tlrst to arri ve. Hence it is looked to as a primary source of high-grade inte lligence. (2) Much of the intercept material consists of odd pieces of information and sometimes no more than a clue, which can be understood or followed up only in the li ght of infonnation that must be obtained from other sources. He nce the processors of intercept intelli gence tend to draw into themselves-and must draw into themselves to be effective-the intelligence output of other agencies . This they must put together with what they have; and if their own information is good they are likely to come out with more accurate and better evaluated information than is elsewhere available. We see thi s constantly in comparing o ur information on va rious topics with infonmtion put out by the B.E.W. , O.S.S., J.I.e. and other agencies.

(3) Because they mu st constantly seek information from outside sources, those who deal with intercept intelli gence have an opportunity to learn what the good sources of information are on a wide variety of subjects. In view of the multiplicity of inte lli gence agencies and wo uld -be intelli gence agencies in Washington , the large amount of duplication of work, and the penchant of certain agencies to bring forth a stream of class ified report s that are inaccurate and, in many instances, not much more than irresponsible guesswork , thi s is a ve ry important point. (For instance, though the Joint Intelligence Committee is among the least of the si nners in thi s respect, I think it fair to say that its Weekl y Summaries in the past two months have contained a numbe r of inaccuracies and some positively e rroneous or misleading information , which thi s Branch , at least, would never have put out unde r an y circumstances.) c. The guiding consideration in the internal arran gements of the organiza-

tion that handl es cryptanalytic mate rial , and in its methods of dealing with outside persons and agencies, is the conside ration of security. One lapse of securit y is all that is necessary to dry up a radio inte rcept source. Therefore, both on the officer level and below, only persons of the greatest good sense and discretion should be e mployed on thi s work. This conside ration is basic, since intercept informati on in volves a different kind of secrecy than does most other classified information. It wi ll make no difference a year from now how much the enemy knows about Ollr

present troop dispositions. about the present whereabouts of o ur nava l forces or about other simi lar facts that now are closely guarded secrets. But it will make a lot of di ffere nce o ne year from now-and possibly man y yea rs from nowwhether the e nemy has leall1ed that in Ap ril 1942 we were reading hi s most secret codes. Not present secrecy. not merely secrecy until the battle is over. but perm a-

ACCEPTING TH E C HALLENGE

57

nent secrecy of thi s operation is what we should stri ve for. That may be too much to hope for in a democracy; but if it is poss ible to atrain o r approach it we should try to do so. d. An essenti al job of the Spec ial Branch, which must be done if it is to produce this max imum usable intelligence, is 10 establi sh intercept pri orities, keep them constantl y in tune with intelligence requireme nts and see to it that those who ope rate the inte rcept stati ons shall understand and fo llow them. Onl y a small fraction of the wo rld 's radio traffi c can be liste ned to, with the fac ilities that are at hand or obtainable. To deal with the priorit y problem there must be a continuous appraisa l of mate rials coming from the various circ uits, of the proba ble inte lligence value of what mi ght come from othe r c irc uits a nd o f the week to week necessities from an intell igence angle. The who le question is infinitely complicated and time-consuming, as witness the fac t that yo u and Col. Tay lor and I spent many days on the last priorit y directi ve. But it is an absolutely essent ial job 10 be done. With Col. Taylor going 10 London we will have to fi nd some othe r man to spend a good part of hi s time on the inte rcept priorit y problem. e. Anothe r essenti a l job is to esta bli sh cryptanal ytic priorities and keep track of c ryptanalyti c projects. Cryptana lys ts are scarce and their time must be used to best advantage. This requires a degree of liaison that we have not yet been able to reali ze, from lack of man-power. Thi s pro blem, moreove r, is closely inte rwoven with the inte rcept priorit y pro ble m. If you are 10 put the cryptanalysts on a new project when they have fini shed the one that they are now wo rking on, you must devote a part of yo ur intercept fac ilities to the building up of a large volume o f traffi c in the as ye t unreadable code, so that whe n the time comes to tackle it the cryptanalysts may have enough materials to do the job. f. Still anothe r essential function of the Special Branch is to suppl y the c ryptanalys ts wi th lead s and clues, and with inte lligence in gene ral. The art of cryptography has outstripped cryptanalysis. to a point where high leve l codes and ciphe rs are no longer broke n by cryptanalytic method s alone, even though modem mac hinery is capable of almost mi rac ulous pe rfo rmance in the anal ysis and synthesis o f cryptographs. Today it is collate ral intelli gence which solves hi gh-grade c rypt og raphi c sys te m s. Whil e th at int e lli ge nce ca n co me fro m a numbe r o f sources, in the nature of things the best source is the inte lligence orga ni zati on that struggles day by day with the cryptanalytic output and is familiar with what the cryptanalys ts are trying to do and what their needs are. From the long term standpoint the Special Branch's function as a feede r o f intelligence to Arlington Hall is hard ly less import ant than its function of getting intelligence out o f what Arlington Hall produces. We have done a lor o f wo rk along thi s line; but we wo uld be doing a great deal more if we had more help . g. The ex pe rie nce o f the Spec ial Branch has proved that astute peopl e worki ng on the spec ial mate rials have opportunities to furni sh inte lli gence th at leads di rectl y to the so lutio n of codes. LI. Littl efi eld . for example, is directl y

SIGNALS INTELLIGENCE

58

responsible for the recent solution of the cryptographic system used by the Gennans in communicating from Berlin to agents in South Africa. From his study of other intercept material he came to the conclusion that a certain class of unreadable radio traffic was being beamed to South Africa, although the intercept people, including those of the Coast Guard, were pos itive that it was being beamed to South America. He insi sted on this position, and informed the cryptanalysts what he thought the messages might deal with; and they applied probable meanings to the messages that they had. The solution of the code followed very quickly, and the contents of the messages proved that Lt. Littlefield's guesses had been correct. h. Intercept coverage and cryptanalytic work has one very important purpose that is sometimes lost sight of, in addition to the job of getting intelligence out of the messages. That purpose might be described as protective monitoring. This has many phases, of which the following are illustrations: (I) We get from S.S.D. a large volume of samplings of [text withheld] traffic, some of which has to do with ship movements, deliveries of airplanes and munitions, assignments of military personnel , placing of orders in the United States, and other subjects having military aspects. Almost none of thi s material is ever reported by us. Nevertheless we follow it very closely, in order to make sure that it does not become a source of valuable infonnation to the enemy... [text withheld] It is our practice to check each item of infonnation to detennine whether it is one that the enemy could not get from other sources, whether, if it is, the infom13tion is of importance and, if it is important , whether the degree of importance warrants taking some action. Generally our conclusions have been negative; but in one or two instances General Strong's attention has been called to such a problem and he has alerted those dealing with [text withheldl to the dangers involved in giving certain information and in doing business in such mrumer as to require too much use of the radio. (2) Another aspect of " protective monitoring" is to keep on the alert for any evidence that the enemy is reading our own codes or the codes of other nations. Since everybody appreciates that codes will be changed or abandoned if the use rs discover that other countries are reading them-and since. therefore, everybody tries to be very careful about transmitting information from intercept sources- the most meticuluous "detective work" is required in dealing with this problem.

*** (3) Still another aspect of thi s subject is to keep track of th e enemy's sources of infonm.tion and to keep posted about what he knows. Prior to the landings in North Africa. attention was directed to all available traffic that might revea l the enemy's knowledge, and it was examined with great care with that in mind. The cutting down of Axi s code communications by Chile and later by Argentina was brought about by our State Department because they knew from

59

ACCEPTI NG TH E C HALLENGE

inte rcepts what kind o f in fo rmation the e nemy was getting. knew how much pressure on the two countries was justified by the fact s, and knew, finall y, what ace they had in the hole in case they sho uld be fo rced to use it. The location by [text wi thheld] intercept acti vities, o f enemy agents who we re sending valuable infonnati on to Be rlin and Ro me, has enabl ed the m to deal with the proble m in some instances in a diffe rent way, viz. , by buying up the agents and getting control over the Axis source. The re a re several places in the world today whe re the problems of Ax is communications arc a li ve and import ant issue. i. U. S. o pe rati ons in the fi eld o f cry ptanalytic inte lligence are re lative ly limited. For various reasons, including the pie ty o f the State De partment and the fact that our Arm y was starved for fund s pri or to the war, we ente red the war with re lati vely little expe rie nce in c ryptanalytic matters, and with even less ex pe rie nce in dealing with intelligence from inte rcept sources. Thi s is certai nl y true as compared with the Bri tish, and we have reason to suppose that the Gemlans- and to sus pec t th at the Ru ss ians-we re also fa r ahead o f us. Whil e we have mad e progress, we have a long way to go. Large as it is, our S.S .O. operation is on a

much smaller scale than the various comparabl e operati ons o f the Briti sh. Thi s is also true o f the intelligence side of the picture; and so long as it remains true, we will continue to be at a di sadvantage

*** m. The re is one aspect of our relati ons with the Brit ish in intelligence matte rs that has struck me ve ry forcefull y. O ur intelli gence agencies in Washington appear to spe nd a lot o f their time keeping infonn ation secre t from one anothe r, in competing fo r credit and in beating one anothe r into print. All of them have some li aiso n or othe r with the British intelligence agencies; and fo r some reason they are freer in giving information to the Briti sh than to one another. The Briti sh do not keep their infomlati on in water-tight compa rtments, nor do their va ri ous age ncies seem to care who gets the credit. They arc so set up that their intelligence is put togethe r and evaluat ed; and the j o b is do ne promptl y. Thi s g ives the m a great ad va nt age over us in intelligence matte rs. n. We ha ve had enoug h to do with the Briti sh to have d iscovered that there is a considerable amount of intelligence which they withho ld from us, not espec ially by design. but because we do not ask fo r it. Sometimes it is not asked for because nobody knows it ex ists. Recentl y. for example, we have d iscovered that certain inte rcept informati on about Cont ine ntal Europe, which does not fl ow regularly into any American c hannels, is available to us at the Embassy, me rely because we deal with cryptanalytic material and have establi shed a certain amount or confi de nce in o ur j udgeme nt and discreti on. If Major He neman , who has pursued thi s contact, we re not encumbered with a large burde n of other wo rk, and could spe nd half his time at the Embassy, I am confident that he could bring into the United

60

SIGNALS INTELLIGENCE

States channels of intelligence a substantial amount of valuable inFormation that either does not get into those channels at present or gets into them too late to be of the greatest value. o. A Fairly important problem that we have--one that will grow with the volume of traffic-is the proble m of increas ing the proportion of Arlington Hall inte rcepts which are at least scanned For their intelligence content. The problem exists in all languages covered, but it is pallicularly acute in the Japanese. Because of the shol1age of cryptanalysts and translators, not all decipherable messages can be deciphered. and not all that are deciphered can be translated. There is a scanning and selecting process at both the cryptanalytic and the translation stages. This is done by the cryptanalysts and translators themselves; and on the whole they do a good job. However, to insure that no useful intelligence is missed, the organization handling the intelligence should participate in the process of selection and should continually educate the cryptanalysts and translators in the week to week intelligence needs. This is done in the British Black Chamber organization, and it is done in the Navy organization. We have assigned Capt. Ervin as liaison with Arlington Hall , and he will be able to do this to some extent. However, there are so many liaison req uirements that an additional officer will have to be provided before long, if we want to be sure that the work is done thoroughly. In addition, there is a spec ific need For a man familiar with the Japanese language, to deal with Japanese translators. We have such a man , Dr. Creel, who because he is also a Chinese scholar can help the translators on one of their most diFFicult problems, which is to identify Chinese names from their Japanese Kana versions. He has already helped them a great deal. But, until we get some additional personnel, he cannot take much time away from the day to day job of handling Far Eastern material. p. The WT/ I field, and the relations of the Special Branch and Arlington Hall 's work in this field , are matters that are now under study, and I mention the su bject only for the record.WT/I-the derivation of intelligence from radio traffic by all means short of cryptanalysis- [text withheld] It has, of course, numerous applications in the field; and it is of special importance to the Navy. Arlington Hall has been experimenting with Japanese Army traffic from the WT/I angle: and the prospects For deriving intelligence from it look promising. But the developme nt of thi s work, to judge from what I have seen of it thus far, is going to require a close liaison with the military intelligence and, from our standpoint. the Full time of at least one man .... I. Personnel Silllolion oj Ihe Special Branch-Ojficer.\· and Civilians. In the foregoing somewhat discursive remarks. which by no means cover the whole subject, I have tried to show that our problems have many sides, that the field of cryptanalytic intelligence is important, and that there are great opportunities in this field if we are prepared to seize them. Whether or not the objectives that we have in mind are sound, I think the problems at least are of some real importance to the country. The adage that knowledge

61

ACCEPTING THE CHALLENGE

is power is nowhere truer than in the field of international relations, especially during a war and in preparation for peace . We are gaining knowledge and experience in many parts of the world, some of it rather expensively; but we still have a lot to learn. In the field of intercept intelligence we have at hand, in my opinion, one of the greatest potential sources of information; but we are not making the most of it. We need to raise our sights. The above statement of our problems may indicate why the present manpower of the Special Branch is not adequate to do the job that we want to do. A quick pi cture of our manpower (excluding you and myself) may be given as follows: Total officers Civilians comparable to officers

30

-.2 39

Deductions must be taken for: Officers who work in the Cryptographic Security Section and Radio Communications Section (transferred from c.I.G.) who have nothing to do with S.S.D. material Officers who do purely administrative work, which is very heavy because of security requirements and the large volume of paper work that passes over your desk Officers about to be transferred to Ground Forces (2) and to London ( I) Mr. Bingham , who has enlisted and is about to be inducted in the Army

5

3 3

I

.ll Balance To get the net working forces on the intelligence job, certain additional deductions of men on special assignments must be taken, as follows: Major McKee (Order of Battle specialist) Capt. Berle (Liaison with the State Dept.) Capt. Calfee (Only technically assigned to the Branch. doing most of his work for AIS ... .) Capt. Ervin (in training at Arlington Hall, to be a liaison officer with

27

62

SIGNALS INTELLIGENC E

that organi zation) Mr. Randolph (about to commence tra ining at Arlington Hall , 10 wo rk on WT/ 1 material ) LI. McCluney (temporarily spe nding full time on personne l work) LI. Wilkinson Uust relieved as Security Offi cer and engaged in training hi s s uccessor)

7

Present net workin g force on inte lli ge nce material Accordin g to the ir work and the parts of these 20 me n are ass igned as follows. the world that they keep track o f: Ed iting " Magic Summary" and studies 3 Read ing and routing collate ra l inte lligence material from regular sources

1

Biographi cal Secti on Far East

2 6

Ncar East

I

I

Easte m Euro pe Central and Weste rn Europe Southern Europe and Africa Latin Ame rica

3 2

1

Conside ring the volume o f materia l tha t we have to handle and the job that we a re trying to do on it , the fi g ures above speak for themselves.

*** A Ifred McConmack, Colone l, G.S .c.

ACC EPTING TH E C HALLENGE

63

From SRH 141 -2 . In a moment of discouragement about the Army's relatil'e lack of success in exploiting Japanese military (as opposed to diplomatic) communications, McCormack wrote this memorandum to his immediat e superior. Like a great many of his memoranda . it dealt with personnel. The B Branch of MIS , mentioned in the correspondence, handled Japan ese military decrypts; A Bran ch handled diplomatic materials; C Bran ch , th e German intercepts provided by the British. Ironically enough, at the time he wrote, SSA and Special Branch were on the verge of a definitiv e breakthrough in deriving intelligence fi'om Japanese Army messages.

3 Jan 44. MEMO RA ND UM FOR COLONEL CLARK E: Subject: Pe rsonne l Proble m of B Section, Special Branch. I. This atte mpt s to state the problem o r getting intelli gence o ut o r the Japanese Amly material, specifically from the standpo int o r personnel require ments. 2. The Unit ed States e nt e red the wa r co mpl ete ly unpre pared to dea l w ith Japanese Ann y cryptography; we we re nOt the n reading, and had never seri ously attempted to read, any hi gh-leve l Japanese Army ciphe r system. (For that matte r we had never tried to read the Germ an military systems e ithe r). 3. T he first break into a hig h leve l Ja panese sys te m occ urred so me time in March o r April 43, whe n the Wireless Experimental Cente r at New Delhi got into the [tex t withheld I wate r transport) system. By that time eno ugh was known about the high level syste ms o r the Japanese Ann y to permi t a dec ision as to the orde r in which they should be wo rked on and the di vision o f the work among the 4 large crypta nalytic uni ts (Arlington Hall , the GC&CS in Eng land , the WEC in Indi a and the Cent ral Bureau at Brisbane). Arlingto n Hall by ag reeme nt was made the d irecting cente r or Japa nese Ann y cryptanalyti c acti vities ; and in d iv iding up the wo rk it ass umed the primary burden or the bookbreaking job on [text wi thheld] 4. As of April the re was no Ame ri can ex pe ri e nce in inte lli gence wo rk o n Japanese high level decodes. In a stud y written up in that month we expressed confide nce in the ability of the Spec ial Branch to hand le the material that was ex pected to now in aft er a couple of months, though we did say that a transfe r of pe rsonnel from othe r work wo uld be required, and that our othe r wo rk would be impaired. unl ess personnel could be recruited promptly and trained. 5. Not until the middle of June did we get approval fo r additional pe rsonnel. By that time our wo rk had ex panded far beyo nd the new all otment o f 20 offi cers; we we re g iven additi onal jobs to do, faster than new pe rsonnel could be rec ruited; and we we re not able to put adequate personnel on the JA work .

64

SIGNALS INTELLIG ENCE

6. Hav ing no ex perience to guide us, both Arlington Hall and ourselves had under-estimated the difficulties that the Japanese systems would present : specifically (it now seems to me), both had unde restimated the amount of intelligence and research work that wo uld have to be done in orde r to build up the wate r transport codebook and get intell igence o ut of the messages. Both organi zations, the refore, were unpre pared for the job thai thi s mate rial presented. 7. By June, it was clear that, in orde r to capitali ze on the opportunities presented by [text withheld] and other syste ms then in process of solution , Arlington Hall wo uld ha ve to increase its pe rsonne l by a minimum o f 2,000 people. Studies designed to present the problem to the appropriate authorities we re commenced. By August a request fo r the necessary pe rsonne l, acco mpanied by a T/O and manning tables, had been submitted. It a ppeared on the face of the stud y that a critical situati on ex isted. Ne ve rtheless, 4 months passed without effecti ve ac ti on be ing taken: and onl y within the last few days has Arlington Hall received the authorit y necessary to commence recruitin g.

8. In other wo rds, 6 months after a c ritical need for at least 2,000 pe rsons was obv io us to anyo ne familiar with the prob lem, the le ngth y and diffi c ult j ob o f recruiting such pe rsonnel is just being started. 9. Meanwhile, upwards o f 200,000 Ja panese Arm y messages (in [tex t withheld] and othe r partiall y solved systems) ha ve piled up on the she lves at Arlington Hall , unread. These messages wo uld be readable if the necessary personnel we re availabl e. Most o f them will never be read , because the priority of current traffi c, the inc rease of readable current materi al at a rate fas ter than the recruitment o f personnel, and the sheer necessity of mo ving fo rwa rd from one pe riod of a system to anothe r, in order to preserve cryptographic continuit y, will make it imposs ible ever to all ocate pe rsonnel to decipher and translate a ny substamial portion of these messages. 10. How much pri celess intell igence is locked up in these unex plo ited and unex pl oitable messages is an ybody's guess. It is anybody's guess, al so, how much info nnation of ope rati onal importance has been lost, because when the opportunit y came along we were not ready for it.

*** 77. O ne thin g is sure. If eve ry limit ati o n of pe rso nn e l a llotm e nt s we re removed , and the Spec ial Branch we re all owed to recruit (as the Briti sh GC&CS do) qualified pe rsonnel from whe re ve r they might be found (offi cers, enlisted men and women and ci vilians), it would still be imposs ible, even within the next yea r, to rind and train the number of qualified pe rsons necessary to keep up with the ever-increasing opportunities which the Japanese Anny traffi c will probably present. 78 . The write r would sum up the situati on presented by the Japanese Arm y tralTic as foll ows:

ACCEPTING THE CHALLENGE

65

a. In o ur exploitation of this fi eld we are far behind whe re we should be, or where we might have been had the job been seen in it s true proportions and been g iven the necessary priorities in manpower and facilities. b. We have arrived, nevertheless, at a point where an opportunity is presented to inc rease our Japanese intelligence a thousandfo ld. c. If we ac t vigorously, giving thi s job the kind of priority that the Briti sh ha ve given their work on Gemlan "E," we have the chance of getting into Japanese AmlY cryptography so deeply that we can never be di slodged from it by changes in systems or by dev ices employed by the enemy. d. If we act vigorously, at the very least we can open up sources of intelligence that will certainly be of great importance, and mi ght even be of deci sive impo rtance, in the Pacific war. e. Except for the matter of I.B .M. machines at Arlington Hall , the problem reduces itself almost who ll y to one of obtaining qualified personnel , when they are needed and not 6 months later. f. If we go on the way we have gone to date, we shall continue to miss opportun ity afte r opportunity, and at the e nd we shall have paid for the lost opportunities in li ves and wasted resources. 79. The foregoing appears to confirm the opinion that you have expressed, that Arlington HaJJ's increase of personnel should be rai sed from 2,000 to 5,000, amd that it should be allowed to go forward to recruit 5,000 qualified person s as promptly as poss ible . When the Wireless Experimenta l Center in Indi a, with a much more limit ed mi ssion , already employs 5,000 personnel and is moving to increase that numbe r as rapidly as personnel can be found, an increase of at least 5,000 pe rsons for Arlington Hall would seem to be indicated. 80. An increase of pe rsonnel for B Section o f the Special Branc h from 55 to 280 officers and from 22 to 120 enlisted women or eq uivalent personnel is a drop in the bucket alongside of the Arlington Hall req uireme nts. It is, moreover, no over-estimate of the requirements, if the Briti sh organization is to be take n as a model. Hut 3 at B.P., an intelli gence section dealing wi th high-level German traffic solely from the cu rrent operational standpo int , had 328 persons working in May; and this is exclusive of administrative personnel. In the two principal sections dealing with the longer-tenn aspec ts of Gemlan military traffic there were 257 persons and 179 persons (agai n excl usive of adm ini strati ve personnel), or a total in the 3 operations of 764 persons. This total is excl usive of a Tabulating Section of 239 persons, and takes no account of personnel on the intelli gence side of the operation at the War Office a nd the Air Ministry. The Special Branch must perfonn not onl y the functions performed at B.P. but a number of those perfomled by the Service Ministries in London. 81. This silLlation is a challenge to the War Department to tack le a new kind of job, one not easy to understand and not easy to fit into a fam iliar pattem of organ ization. and one that requires considerable imagination and a particular type of personnel.

66

SIGNALS INTELLIGENCE

82. It is also an intellectual challenge to the United States Command. The victory of the Briti sh over the Gennans in the radio intelligence field is a great intellectual victory of one leader and one people over another. The Germans have been outwitted at every

tum ~

their most secret military communications are read every

day; the ir most elaborate security precautions count for nothing. Wh y were the British able to get to the "Scharnhorst" and si nk her? Simply because they were able to arrange their tactical di spositions and deception measures in the light of knowledge, derived from reading messages in what the Gennans st ill consider an unbreakable cipher, of the vessel's precise mission, the nature of its escort and the exact time when the battle convoy was go ing to leave Lopphovet. 83 . Except for whatever our Navy has been able to do, we have ac hieved no corresponding victory over the Japanese. To date, in the fi eld of radio intelligence, we have met largely with failure . 84. The chances for victory, however, lie before us; the o nl y question is whether we have the vision to see them and the imagination and energy necessary to seize them.

Alfred McConnack, Colonel, General Staff Deputy Chief, Special Branch, MID

CHAPTER IV Training the Force The very large si gnal s intelligence effort mounted by the U.S . Army in World War II demanded masses of trained personne l. For most of Wo rld War n, officer and enli sted crypto logic personnel were trained at Vint Hill Farms in Warrenton , Virginia, which was al so the site of the Sig nal Sec urity Agency 's Monitoring St ati on Number I. The "Official Sto ry" of offi cer training is contained in the excerpt from SRH 358, Hi story of the Training Di vision, Si gnal Security Agency. A more info nllal account of the human side o f the training process as it affected enli sted ranks can be found in the 'T hird Battle of Manassas," an unpubli shed manuscript prepared by some of the trainees.

68

SIGNALS INTELLIGENCE

From SRH 358. An excerpt from the History of Training Division , Signal

SeCllrity Agency, discusses the training of cryptologic officers at Vint Hill Farms Station during World War II . Indicative of the fa ct that the Agency (and the Army) ref/ ected all too well the prejudices of the time is the fa ct that all officers selected 110t only had to be mentally and physically fit , but also "of the white race."

CHAPTER IV HISTORY OF SCHOOLING OF COMMISSI ONED OFFICERS IN THE IIINT HILL FARMS SCHOOL \liNT HILL FARMS STATION WARRENTON, IIIRGINIA , 17 M arch 194 1 to I Jalluary 1945

*** 10 March 1942 10 5 October 1942 On 10 March 1942, the training of commi ssioned o fficers was initiated in the CryplOgraphic Division . Fort Monmo uth , Ne w Je rsey. The Cryptographic Di vision was the n a part of and unde r the supe rvi sion of the Enlisted Men's De partment . Signal Corps Sc hool, Fort Monmouth , Ne w Jersey, and had been organized prim aril y to train enli sted men in cryptanalytics and lingui stics. The training of commi ssioned o ffi cers was initiated prima rily because of the need in various Signal Corps and Arm y Air Fo rces units and in stallations fo r pe rsonne l qualifi ed as Message Celller Officers, Cryptographic. MOS 0224. It was certain that e mphasis in such tra ining should be placed on c ryptography and , the refore. that a school organized to teach cryptanal ysis would appare ntl y not be the prope r place to g ive such instruction. However. the Cryptographic Di vision was the onl y Signal Corps school then ex istelll whi ch could not onl y teach the ope rational phases of War Department syste ms but could also demonstrate to the student s by crypt anal ytic me thods. the ir weaknesses and thus e mphasize to the m their responsibility for the security o f such systems. On 10 Marc h 1942, a group of sevellleen officers. assigned to the Signal Corps. were placed in suc h training and fift een we re graduated after a pe riod of a pprox imate ly te n weeks. All of the gro up graduated were transferred to the Arm y Air Forces for c ry ptog raphi c sec urit y duties. Additi onal o fficers we re ass igned for training from both the Signal Corps and the Arm y Air Forces from that date until 5 October 1942, so that the ave rage numbe r graduated pe r month during that peri od amoullled to te n, o f which approximately 50 pe r ce nt we re ass igned to dut y with the Arm y Air Forces and the remainde r to dut y with the Si gnal Corps.

TRA INING THE FORCE

69

Since the commissioned training durin g thi s period dea lt with relatively small classes, only two commissioned instructors were needed and these were provided from officers assigned 10 the Staff and Faculty, Signal Corps School. The instructors were required to be trained as Cryptanalytic Officers, General, MaS 9600, with spec ial additional training in the operation and analysi s of War Department systems. All commissioned students were required to have an AGCT grade of 130 or better, an analytical mind, and preferably a college education. All had to be cleared under the provisions of classified AGO le tte rs pertaining to the c learance of personne l for cryptographic duties. Due to the fact that almost all position vacancies we re in overseas units or installati o ns, composed of or operated by white troops. all officers had to be of the white race and physically qualified for overseas service. The problem s encountered in thi s phase of the trainin g conduc ted by th e Cryptographic Divi sion were relative ly few and based primarily o n the difficulty in obtaining current and pertinent instru ctional material and on the diffi c ulties created in the scheduling of training o n classi fi ed material by reason o f the delays inc urred in Obtaining cryptographic clearances. As the student body increased , the further problem of finding adequate classroom space was c reated. No additional space was available a t Fort Monmouth, and as a conseque nce, the school was placed o n a two s hift basi s. No reports have been received critic izi ng the adeq uacy of the training g iven commi ssioned officers durin g thi s period , but it can be prope rly acknowledged that by reason of the speed-up created by the demands for trained o ffi cers. the training was at best weak because the theoreti cal and technical phases had to be e mphasized at the expense of practical application, and as in most types of school training given to commissioned officers, little or no opportunity was a ll owed to them to gai n experi ence in the ad ministration and hand ling of troops. Enli sted overhead for the Cryptographic Divi sion durin g thi s period we re assigned to th e Second Signal Service Company, Was hin g to n , D.C. and o n detached service with the 15th Signal Service Reg iment , Fort Monmouth, N.J.

5 Ocrober J942

10

dale

During the s umme r of 1942, negotiations were initiated to dete rmine the feasibility of moving the Cryptographic Divi sion to some locati o n near Washington. D.C. Thi s was deemed advisable in order that closer liai son could be obtained with the Signal Securit y Agency, so that more c urrent cryptanal ytic info rmation could be obtained, and so that more physical space would be available for both commissioned and enli sted training. On 5 October 1942, pursuant to authority contained in paragraph 2. S.O. #272, Hq . Fort Monmouth, N. J .. dated 4 October 1942, all overhead and student person-

70

SIGNALS INTELLIGENCE

ne l the n assigned for duty to the Cryptographic Divi sion were transferred from Fort Monmouth to Vint Hill Farms Station , Warrenton, Virginia, a Class IV installation . At Vint Hill Farms Station. the name of the school was changed to the Si gnal Corps Cryptographic School. The transfer of the sc hool , although well intended, was unfortunate at the time, due to the fact that Vint Hill Farms Station was the n only in the initial stages of construction and certainly not, at that time, a proper place for the operation of a special service school. Due to the size of the stude nt body and the initial lack of classroom facilities, it was necessary, unti l all of the present school buildings were completed in April of 1943, to operate the school on two shifts, which necessarily resulted in c utting do wn the effectiveness of the instruction give n. However, once the physical facilities were completed they were e ntirely adequate except in rare cases when it has been necessary to turn do wn unanti cipated training demands. During the remaining part of 1942 and during 1943. the courses of instruction given to commiss ioned officers were increased to qualify officers in the followin g military occupational specialties: MOS 0224 Message Center Officer, Cryptographic Cryptanalytic Officer, General MOS 9600 Cryptanalytic Offi cer, Translator MOS 9604 Cryptanalytic Officer, Traffic Analyst MOS 9605 In January of 1944, a section was added to the school to instruct officers (as well as enlisted men) in the installation , maintenance and repa ir of cryptogra phic equipment and to qualify officers with the foll owing military occupational specialty : Cryptographic Equipment Mainte nance & Repair Officer MOS 9600. In May 1944, pursuant to directive of the Chief Signal Officer, instruction of Message Center Officers, Cryptographic, MOS 0224, was made the responsibility of the Eastern Signa l Corps School , Fort Monmouth, N. J., and was discontinued by the Signal Corps Cryptographic School. In June 1944, it was recommended by the Chief Signal Officer and approved by the Commanding General, ASF, that the school he nceforth be known as the Vint Hill Fanns School. During the period under discussion, all commissioned overhead for the schoo l have been assigned to and furnished by the Signal Securit y Agency, and all e nli sted overhead ha ve been assigned to and furnished by Detachment , Second Signal Service Battalion , Vint Hill Farms Station. No Table of Organi zation has been created for this purpose, but commi ssioned g rades and e nli sted grades and ratings ha ve been allotted , suc h allotment being re lated directly to the training demands and thus increasing and decreas ing as the training demands increased or decreased. This has resulted in a desi rable fl ex ibility of overhead and has proven entirely adequate to serve the training demands of the school. All overhead, though being furni shed by the organizations indicated. have with few exceptions been graduates of the Vint Hill Fanns School. All commissioned overhead wi th few exceptions ha ve

TRAINING TH E FORCE

71

been eithe r enli sted students or ove rhead in the Vint Hill Farms School pri or to their being commissioned . The primary proble ms in all the courses g ive n by the Vim Hill Farms School have been to provide sufficient pe rsonnel to meet theater demands and to keep the courses current in a constantl y changing fi eld . The first of these two problems is an unso lved one. Altho ugh it is believed that suffi cient commissioned pe rsonne l have been trained to me t theate r demands prior to this date, it has been primaril y a res ult of fortunate assignment of offi cers to the courses given . At no time has adequate informati on been forthcoming from hig he r headquarters on whi ch a reasona bl y acc urate estimate of future training demands could be based. When it is considered that all of the courses give n by the Vint Hill Farms School must diffe r in the fundame ntals the reof, de pendent upon the theater in which the student is to be assigned, the problem in making a reasonably acc urate estimate is self-evide nt, but by the same toke n, it is also self-ev ide nt that a bad guess wo uld result in an excess of officers being trained fo r one theater and an insufficient num ber trained fo r another. The pro ble m of keeping the courses current has been met by constant and c lose liaison with Signal Securit y Agency and by requesting through the Signal Securit y Agency, reports from the theaters pe rtaining to the adequacy or inadequacy o f the training being given. The methods of instructi on have va ried with the diffe rent courses. Orig inally, whe n adequate time was av ail a ble, emphas is was placed o n individu al stud y. However, as demands fo r commissioned personne l inc reased, it was necessary to speed up the courses. Lectures, demonstrations and indi vidual and team solution have been the principa l method s of instruction used in cryptanal yti c and traffi c analytic courses. Lectures and inde pe nde nt stud y have been the principal methods of instructi on used in language subjects. Lectures, demo nstrations, inde pendent study and coac h-and-pupil-method have been the principal methods of instructi on used in the cryptographic equipment mainte na nce courses. Adequate commiss ioned stude nt pe rsonne l have been ava il abl e fo r training and with the exception o f those officers on detached service from the ir units. all s tu de nt offi ce rs have bee n a ss ig ne d to th e OS C RP, Ad van ced Radi o Commun ications School , Arlington Hall Station, Virg inia and pl aced by such organi zation on temporary dut y at the Vint Hill Fanns Schoo l. The mental qualifications and educational background of all offi cers sent to the Vint Hill Farms School for train ing have been adequate. In gene ral, they have been required to have an AGCT grade in G roups I or II if college graduates, and in Gro up I if high school grad uates. Also, all o ffi cers were requi red to possess analytical abil ity. With few exceptions, drop-o ut s have been caused by the lack o f analy tical ability. an abilit y which cannot be readil y pre-dete rmined but will reveal or fai l to reveal itse lf onl y during the course of instruction. With the exception of the normal field and techn ical manuals used fo r general military training, all trai ning has been based on special texts, documents, devices,

72

SI G NALS INTELLIGENCE

c harts, mock-ups, etc. , pre pared e ithe r by S ig na l Securit y Agency o r by the instructi ona l overhead of the school. This situati o n has been nece ssitated by the fact that a lmost a ll technica l instruc ti o n is o f hi g hl y classifie d nature and th a t it is no t intended that inFormati o n pe rtinent the reto be ava ila ble to the aml Y at large .

••• The present o rganizational status of the vint Hill Fanns School is that it is a spec ial service schoo l ope rated under the directi o n of the C hief Sig nal O ffi cer fo r training pe rsonne l of the arm y in cryptanalyti c, traFfic anal yti c and c ryptog raphi c equipme nt mainte nance subjects. The Commanding Officer, vint Hill Farms Stati o n is a lso Commandant o f the vint Hill Farm s Sc hool. The Director of Training of vint Hill Fa rm s Stati o n is directl y in c ha rge o f trainin g in th e school. The school receives technical inFo rnlation and a id fro m the Sig na l Security Agency, and the Commandant of the school is res po nsible to the C hie f Sig nal Officer, thro ug h the Command ing Officer, Si g na l Securit y Age ncy, fo r the adequacy o f the tra in ing conducted .

••* Fro m Th e Third Baule of Manassas. This highly unofficial history of the

training of Army en listed signals intelligence specialists conveys a good sense of th e fla vor of trainin g camp life ill World War II . Lambro s D . Callimahos, the "flute-playing, Esperanto-speaking, Egyptian-born Greek" lieutenant mentioned in the text, went on to a distin guished professional career in the National Security Agency. III an introduction he wrote to the original manuscript , Callimahos comm ented that "Sometimes it is a bit catty. sometimes biting in its remarks; nevertheless this history is a lot more honest than other histories I ha ve read labeled ·official' ."

TH E THIRD BAlTLE OF M ANAS SAS

••• TH EY F ELL O UT IN FRO NT O F T HE ORDERLY ROO M At e ig ht o'clock o n the mo rnin g of 5 Octo be r 1942, approx imate ly 250 speculati ve e nli sted me n and th ree o lficers, each w ith a ba lo ney-a s-us ua l box lunch, assem bled a lo ng the Je rsey Central siding in Fort Mo nmo uth , N .J . A coach tra in backed in fro m the directi o n of Eato ntown and sto pped with the baggage car near

73

TRAINING THE FORCE

the general site of the mountainous pile of foot lockers and barracks bags. A detail consisting of about fifty strong " volunteers" was picked to stow the mountain on the train . This done, the men got aboard under the guidance of Lts. J. T. Guernsey, W. P. Bundy, and R. L. Markland and waved goodbye to about thirty fellow students who were left behind, some to pack up the school supplies for forwarding, others to carry the school banner to OCS. During the entire embarkation procedure, the Post Band played familiar airs and cheered the School on to new adventures-or were they just celebrating the departure of the "geniuses" who, in some companies, drew no KP or detail? The destination of the movement was, except to knowing few, only a subject for rumor. Stories were current of Arlington Hall and other comfortable places in or near Washington ; another version had it that the future home of the school would be an old estate in Virginia with a mansion house .... After traveling over the rails of most of the lines in the east and south for about ten hours and finally ending up behind a Southern Railway locomotive with a bright green shell around the boiler, the men and officers arrived in Warrenton, Virginia, where they were met by a diverse collection of trucks and conveyances and most of the Warrenton populace.

*** The trip from Warrenton to the new home was, for some, a sightseeing tour; for those jammed into the front of the trucks, it was a period of slow suffocation inside raincoats. It had rained most of the afternoon and was still drizzling. The sight that greeted the men upon arrival was one of mud and desolation ; and the grim discovery that straddle trenches served for latrines dropped the "campustry" majors to even lower depths. Morning brought a more cheerful outlook, because the sun was bright and the leaves were green. When it was realized that eating or washing required, for most men, a trip across the post, the cheerfulness evaporated again. Although, at the time, few people knew it, Company "B" already existed on paper and included the "Schoolboys," 2S " linguists" of questionable proficiency, and three untutored non-coms, "Lil Abner" Hodges, " Louie" Fragass i. and " Repulsive" Ahearn. These diamonds-in-the-rough were somewhat doubtful as to how to handle the " new fellers" who were "pretty much like ordinary peoplesmart, but kinda mad". Lt. Nixdorff, he of the long chin, was company commander, in a remote sort of way. The " first sergeant" .... had a pat solution for every difficulty: "You' ll have to make your own provision s for that." When Lt. Marion became Company Commander a month later, nobody knew the difference. Company administration was strictly a hit or miss affair. Paper work consisted of a detail roster that included everybody. There was only one copy of this and Jones, who had it, could never be found. As a consequence, the efforts of Hodges and Co. to control goofing off were largely nullified. Reveille, also handled by Li ' l Abner, was an indeterminate proceeding. It was held under blackout conditions

74

SIGNALS INTELLIGENCE

where " Barrack so-and-so all presnnnneraccounted FOR'!! was a mere fonnalityexcept that Ahearn's report from the one-holer generally agreed with the facts. The Orderly Room was a clapboard shack at the main gate that looked like a tool shed (and was). With the detail situation very much out of control . a clerk was needed to get out additional li sts of names. A private named Griffin, who was more at home with a typewriter than a shovel, volunteered, and genuine administration got a definit e, if modest, start. A week or so later, the Orderly Room moved to a leaking tent behind Barracks 10, which at the time was only a pile of prefabrication in the rain. When it was at last put together (at a cost rumored to be $3,030 without steps and screens). the company headquarters moved inside and Lt. Guernsey took over as CO. In less than a month , a telephone had been installed and the Company had established communicat ion with the rest of the Post, provi ng that the Signal Corps could do the job if the customers would just be patient. The tent became a telephone booth outside of which long lines of patrons formed at odd hours of the night and morning.

Shortl y afterwards, a group of officers and " non-com material " ... left to start "C" Company (later Hq. and Hq. Co.), whose original function was basic training for an assortment of illiterates, ex-convi cts, gi llar players and spare parts who had shipped in when nobody was looking. Hodges' bunch moved over, too. In those days the immediate ambition of every student was to get in one day a week of school. The company was growing: classes were fill ed and were run in shifts. Life was mostl y detail-chopping trees, digging post holes, laying sidewalks, and cleaning the (outdoor) latrines. Life centered around those woodsy chapel s: while the dry and shining bowls in the regular latrines remained unblemished, men from "B" and "c" companies shared a common interest in an incredible twenty-holer on the corner ac ross from Barracks I. There they wo uld med itate amid the fragrance, the "e" men eating candy or munching an orange, the "8 " men

reading Tri stan and Isolde or a volume of T. S. Elliot. ... Having been spec iall y se lected at process ing centers because of superior mental and educational qualifications, the men were put to work at digging and mov ing logs. One count sho wed forty three years of college in one ditch (8 EM); the postholes were likewise "educated." Products of the lumbering activities were classifi ed into (a) logs, (b) branches, (c) trimmings. On odd days the log poles and brush wo uld be moved to one side of the area "to make way" for further developments; on even days they would be moved back again "to make way" for still other developments. This went on until all concerned ran out of developments. It was then decided to bum the brush piles and carry o ut the logs. Those who goofed off got a week's KP-which was in many ways a reli ef. ... During this rugged period while construction was still in progress, casualties were not uncommon. The uncharted maze of trenches, later used as ditches in which to lay sewer and water pipes, probably taking the largest toll. Somehow there seemed to be very lilli e cooperation between the engineers digging the

75

TRAINING THE FORCE

trenches and throwing up the breast works and the other combatants on night operations. While on a spec ial mi ssion one wann October night , Affle rbach, clad in GI undies, walked into a thirteen foot ditch but relaxed (or was relaxed) enough to suffe r noth ing wo rse than mortificati on. On the othe r hand . the same ditc h was responsible for a wre nched leg and th ree bro ke n ribs from two othe r victims that same night. ... Afte r that, extreme caution was observed by the personnel of the Company; anything that looked like a dit ch was avo ided. and the dim morning hours found dozens of me n g ingerl y stepping over the shadows of trees. Soc ial life on the Post was somewhat suppressed by the lac k of fac ilities. O ve r in "A" area the re was a barrac ks that had been converted into a day room, mail room, library, spec ial services offi ce and PX . The latter consisted of a twelve-foo t counte r a nd three she lves, mostl y em pty; bes ides , it was nea rl y a mile away. Dinne r invi tati on ex te nded by the Warre nto n fo lk (that was before the nati ve money-grabbers made a mocke ry o f "Southe rn hospita lity") gene rall y had to be decl ined: eleven miles each way was a long wa lk, even fo r the Signal Corps. When Sgt. Parker tried to organi ze an orchestra, the response to first call consisted of five trombonists and two xylophone players. It was seve ral months before a combination with the customary assortment of instrume nts could be gOllen togethe r.

*** Whe n Griffin was away or ill , O wens took reveille with an intonati on that added to the O we ns legend . In HOdges' day, the vocal side o f reve ille had been no thing more exciting tha n a success ion o f strange no ises in a Kentucky twang. Whe n G riffin came along, he enli vened the proceedings with a haughty " re h-PORHT! I " and a batte ry of announcements and involved instructions proclaimed in hi s best theatrica l manner. Against thi s professional competion, Owens staked hi s bid fo r fame o n a n unme lodi o us " RE-E- E-E-E-E-E-- PORT" de li ve red in a dro ning monotone. With the Griffin-Owens team running things, the me n could de pe nd on a reveille in harmony with the di zzy e vents o f the later hours. About this time, Lt. Vo lm , dec iding that it was time for rec reation and beauti fication, began the Area Cleanup Project. Lt. Kun zman became sidewalk eng ineer and used for hi s theory o f construction "straight logs, onl y straig ht logs ." Th is meant that logs wo uld be toted in fo r the wa lk one day and o ut the nex t- fo r longer and straighter ones. Again big brush piles burned in the midd le of the " stad ium." Lts. Guernsey and Vultee would stand about the huge fires wa rming their hands (?), sayi ng " mo re branches on thi s side," and, " more twigs over the re .... let 's gel a move on." Danny Vo lm was a skillful " improv iser. " Whe n the sidewalks needed gra vel and it was not fo rthcoming from the engineers, he wo uld stroll up to the mai n gate, and, as the trucks approached with gravel for the Colonel's wa lks, wo uld sw ing aboard and holle r, " Down this way" and bring them to the company a rea. In thi s

76

SIGNALS INTELLIGENCE

manner several hundred tons of shale were obtained for a 220 yard race track that ran around the area between the barracks and di vided Goldbrick Park in half. It was boycotted by everybody: Afflerbach tried to use it, but recognized the disadvantages of ostracism in lime. Later it was removed , as painfully as it had been built. The Day Room began to shape up unde r Volm technique. Huge slabs of masonite for the new white barracks in "C" company disappeared, only to reappear on the walls of the day-room, neatly squared off with lathes gathered in the same manner. ...

Inspections were frequent. On Saturdays the inspection pany fo llowed a transposed route among the barracks in an allempt to catch the boys warming their hands around the stoves instead of standing by their bunks... .This "hit-em-wherethey-ain 't looking" strategy worked occasionall y, particulary in Barracks I (the goat: Kilgore). Captain Guernsey was a no-haircut g igger, especiall y when the barber shortage was ac ute. Homer Baker got it for "dust in ear" -lint from an 00 towel, it was. Lt. Vultee spec iali zed in upper bunk g igs where you could get an eye- leve l view of the wrinkles in the bottom sheet. Lt. Kun zman who usually inspected with him , however, speciali zed in lower bunk "gigs" which were more in line with hi s qualifications.... Lt. Volm liked to look for food which he ate (except Biern 's beer-cheese), leaving a gig s lip in gratitude. At foot-locker inspections, the barracks leader was required to wriggle around in the airspace above the ceiling vents in a nothing-up-here-but-us-chickens act. All kinds of junk were di scovered: radios, a crate of oranges , satc he ls, volumes of poet ry in limp leath e r, and Wilhe lmy 's six-month fil e of the New York Times. (This practice was ended when somebody fell through the ceiling). Epps was caught with an axe in his footlocker; Jurgens had silverware in his overcoat; Oppenheimer had an aspirin under his bed. Cold days also bred di sease. One after the other, barracks came down with measles or allied bugs that led to bigger quarantines. Finally the whole Post was put in a quarantine based on the following principles : I. That men li ving off the Post were immune. 2. That the gern!s di stingui shed between enlisted and commissioned personnel. 3. That the men ceased to be carri ers every third day but became dangerous again on the fourth (the issuance of passes on Mon.-Thu. -Sun. , etc., was based on this principle). 4. That short periods of immunity occurred in a scheduled manner (this permined operation of the PX on a shift basis). 5.

That contact was dangerous in training activities but haml1ess while on detail.

***

77

TRAINING THE FORC E

ADMINSTRATION FOR ALL CO NCERN ED II wasn' t until three or four days after our a rrival at Vint Hill that the admini stration overhead took up duties in the barn in the now " B" Company area .. .. LI. Guernsey opened hi s offi ce in the rear east room o f the barn . Of course, the school had 10 have a phone so immediately one was supplied (a c rank-type affair). There was a trunk line-to one place, Post Headquarte rs in the bi g house- but to get there it went ac ross the floor of the office, down the hall and o ver the fie lds. If yo u had business with other than Sgts. Enright and Jones yo u had to go for a hike. Afte r all , it was onl y a lillie less than a mile ove r to the orde rl y room at the present site o f the main gate. Bes ides, if it was LI. Ni xdorff yo u were looking for, he wasn' t the re !! One o f the big jobs whic h had to be done again was the roster o f EM in school who were eligible to be T /Ss. Of co urse, we had made a roste r before we left FI. Monmo uth ; what happened to that nobody knew. An yway, on Octo ber 27th the post-ho le diggers got their rank . By that time, it was time to move and the school set up headquarte rs in the tele phone building. except that Reproduction, Suppl y and Pl ans & Training stayed in the (unheated and diml y lighted) allic o f the barn . Those we re the days of easy ratings. Radio personnel we re chang ing stripes as frequentl y as they changed underwear. Then it was leallled that all the overhead were to become at least T ISs, so up to AHS we nt a li st o f men recommended fo r promotion. Everybody was put o n the li st but SNAFU-o nl y about one-half of the men recommended we re promoted. There was a gnas hing of teeth and muc h weeping and wa iling-and no chaplain to issue TS slips. (Sgl. Bruce " I have a few ah-nowns-ments, fel-Iows" Parke r acted as a sort o f chapl ain, but he was busy most o f the time fi guring out schedules for the "Green Dragon"-a conve rted auto traile r which linked the camp with the outside world.) It lOok almost two wee ks to arrange to have the othe r EM in overhead promoted to T/4; afte r that , the re was rank all over the place. During the lalle r part of November the o ffi ce o f the school was moved ac ross the post to the present school suppl y building. Th ree space heate rs (the regular q uota for heating buildings in the Sunny South) were installed. II was hard at first fo r the Warre nton Commandos-LI. G ue rnsey. Sgl. Ka ne, and Sgl. Hawes-IO adjust themselves to the vagari es of the tempe rature in thi s new office. After they stood around the stove for two ho urs, the ic icles wo uld melt and they could slowly wo rk bac k to their desks and get admini stration unde rway. By noon the building would be so hot to stand, while the atmosphe re wo uld be blue from the smoking o f the stoves and the Commandos.

***

78

SIGNALS INTELLlGENCE

... Sunday classes were scheduled on alternate week-ends for all personnel from 0800 to 1645 during January through April 1943. Target practice with a knife was furnished by Lt. Mann in the main office, the target being the then present calendar month. However, instruction was not of the best for he couldn ' t hit the target as often as the student s. In the early part of April it was dec ided to have post headquarters and the school together. And for the fifth time the administrative staff was packing up and bossing a bemused bunch of men who were wandering through the bushes with typewriters, file drawers and in-baskets. Now though, it was clear that progress had been made: for the first time, the school headquarters looked like an offi ce and was uncluttered with dirty stencil s, crates, mess tables and student officers .

*** ... the desk in front of Afflerbach is not occupied but it is reserved. Perplexed student officers spend days at thi s desk filling out multitudes of forn1s , screaming they had a lready fill ed them out at Arlington. After the first da y they become hoarse, the second day they give their occupations for the last five years in answer to an inquiry about their grandparents ' dates of birth , and on the third day they are veteran student officers. Lt. Callimahos is over in the other corner but is heard from all the time. Hi s byline: " Don 't you know how to report to an officer????" .... It was during thi s period that the output of memorandums reached its peak. Although Capt. Guernsey and Lt. Vultee were only four feet apart, they found it necessary to carryon communi cation s by indirect di scourse. The Lt. would jot down notes to be typed by Sgt. Doering who wo uld route them through Sgts . Jurgens and Kane up to the Captain, each Sgt. retaining a copy for file. Memos for general distribution were written by the officers, including Lt. Callimahos, and we re sent down to Reproduction for ditto-ing, while over in the corner Sgt. Kane kept up an envious grumbl e about the days back in Monmouth when he put out all the notices. When the paper shortage began to be felt , economy was effected by substituting buckslips for individual copies. As the practice grew, magaz ines, bulletins, and other printed matter came to be circulated by thi s means, " FYI " and "Note and initial below." Eight y- page copies of Tactical and Technical Trends would be sent out to thirty-odd people in six different buildings " to be returned promptly." Instead, they would di sappear. Sgt. Doering, who used to be a stati stician, got out the ca lculator and some probability tables and estimated that buck-slips issued in May could not, the way things were, be expected back before the middl e of November. Runners were sent out to find the mi ssing buck slips, but no luck. The device of including breezy comment (e.g. "Pages will not be dog-eared") was tried, without success. Finall y, Lt. Vultee blasted out with hi s fam o us memo on " Bucking of Buck Slips" in which negligent " buckees" were charged with retarding the war

79

TRAINING THE FORCE

effort and were warned that "such impedition will not be tolerated. " Everybody reached for his dictionary and awaited the next dispatch . ...

*** ... as Rosh Hashana approached .... Almost the entire student body exhibited religious fervor and a sudden desire to attend serv ices. For thi s, of course. a three-day pass was necessary.

Eq uinoctiall y. these " Passover passes" (the term app lied to both Passover and Yom Kippur) brought on righteous indignation in some quarters because "They get off at Christmas or New Year's, too, and that ain ' t fair' " The other quarters said "Yes" and we nt on pass anyway. ... . . .the Room 3 1 gang .... had started a map in the second stall of the latrine down the hall from the office, a map which included some of the prominent personalities of the period. Irritated (no doubt, by the misuse of government property), Owens issued a notice to "all EM " wa rning them that " anyone found g uilt y of writing on , marking or in any way maliciously defacing any wall , partition or other part of the school buildings will be subject to severe disciplinary action." Though the map grew and grew, Owens couldn ' t catch anybody. As its reputation spread, visitors from other bui ldings and even the Company came over to see the newes t " terrain features ."

Captain Guernsey reg ul arl y passed by the officers ' stall in o rder to keep abreast of developments ....

* ** The Legal Department moved in short ly afterwards and immediately took over. .. .They seemed to regard the school headquarters as intruders. Business was brisk. One day somebody who sa id he was a representati ve of Annour & Co. called up and insisted that Lt. Prall had some chickens for sale at hi s farm. The " Lega l Eag le" ex pl ained that thi s was an anny post, that , anyway, we were using all the chickens we could handle . ...

*** The Chief Clerk has also always had a secondary function- that of a combination John Anthony- Ma Perkins-medicine man for the soul. At least once a day he received a wai ling vis itor with a long story of a check that had to be cashed, a telephone ca ll for a life-or-death matter, bonds that had to be bought , or " I wa nt to see Lt. Kunssssman." The EM long ago learned that the simple story is not always the best. There was . . .the lads who felt all right in the morni ng but now were a little sick, and would like to go to the dispensary, please- passing the PX on the way, of course. Then there were the delinquents sweating o ut an interview with the Captain

80

SIGNALS INTELLIGENCE

or the Exec utive Officer and who stood around for fifteen minutes to a half an hour, gazing aimlessly at the ceiling and gelling out of people 's way while maintaining a pre-execution look of "No, I won't take an eye bandage." And the re were the potential OCS men who had been told to re pon to the Chief Clerk about their applications. They marched down the hall s, sile ntl y counting cadence; executed a sman column left; and, coming to allention at the Chief Clerk's desk, they we re ready to reel off the general orders, the position of the soldier, or a fifty word essay on " Why I want to be an Officer. " Whe n they were asked merely to sign the ir name three times, their relief was so evident they ' d have let out a sc ream (if they hadn 't been OCS men). In addition, the Chief Clerk is often kept quite busy with a security problem peculiar only to School Section "A." Room 36, adjacent to his desk , is the secret classrooms o f all secret classrooms, and the subjects taught there are verbote n to all outsiders. Unfonunately, the human voice carries; and whe n LI. Prall lectures, the windows rallie in the Big House a half mile away. This keeps the Chief Cle rk busy with phone calls from the main gate, as the guards are quite inte rested in his course, and frequent ly request re peat s on cenain paragraphs they mi ssed while checking trucks out of camp .. ..

* **

EDUCATION IN DARKEST VIRGINIA

*** Just at chow time Thursday evening, October 8th, three large commercial vans pulled in to camp with all the worldly possessions of the school that had been left at Monmouth . In addition to the stati oner's supplies and pe rsonal material s of the students, they had all the furniture, twelve tons of IBM paper, the IBM mac hines and a three-ton safe. The draft-dodging that we nt on that night when a detail was needed to unload has never been equalled by any of the " Zoot Suit Clan ." The mate rial s were for the most pan put in the lert of the barn , and the overnow in the firehouse. By eleven PM everything was unloaded but the safe, which was strictl y an "Everybod y-boss-nobod y-do-an ything" job. The si tuation finally got to the point where LI. Bundy and Lt. Mark land gave up and Affierbach first laid c laim to an SSN for sare moving. Sometime during the following week, schoo l was staned with two four hour shifts in the left of the bam, two small rooms on the ground noor, and a barracks. During the fonnative period of the school most of the overhead either drew their materials and continued their studies or were busy pushing details preparing the buildings. Sgt. Frick did neither but instead went to Washington every ni ght and came in just in time for noon chow. . . . As soon as lights were available for evening stud y, a night study hall , optional

TRAINI NG THE FORCE

81

to all stude nts, was in stituted. The roste r for the s upe rvis ion of th e hall was arranged so that one so-ca lled senio r instructor and one junior we re on each night. Thi s probably became as confused as the arm y has ever known , with such arrangements as the fo ll owing prevailing. One ni ght Sgt. Netherton was the pe rson who actually served, but he was doing it for Bailey, who was supposed to ha ve done it for Hurst, who was supposed to re pl ace Affle rbach, who had orig in all y been schedu led for that ni ght. . . . By No vember the school quarters had been moved to what late r became HQ Company's mess hall. The heating arrangement in the mess hall was poor enough with the best of atte ntion , but to make matte rs worse, Cpl. Symmes was put in charge of building the fires. A less apt person for the job could not have been found in all the school. For instance : one cold Sunday, Symmes was the only person in Barrac ks I, and tho ugh hi s bunk was next to the sto ve, he le t the fire go out and went to the latrine to finish his usual Sunday bundle of lette rs.

*** In the compan y by thi s time thin gs had becom e well e nough organi zed to attempt a fire drill and gave the twelve Assistant Fire Marshals some practice. Everythin g we nt ve ry smoothl y until Lt. Mann no ti ced that Sgt. Bail ey had allowed hi s quarantine men to fall out with the rest of the compan y and proceeded to in fo nn Bailey that he was in error. To this he received the philosophical repl y, " Sir, fire recogni zes neithe r race, creed , or co lo r. " Thi s little e pisode may ha ve been the beginning of the Lt. Mann-Sgt. Bailey fe ud which lasted for the duration of Bailey's stay at Vint Hill . On Decembe r 13 school was moved to "E" Bldg. No. I, the fi rst time on the Manassas front that decent quarte rs were available. There now was all ocated one offi cer instructor to each room in addition to the NCO in charge of the room, and in this arrangement , it seemed that Lt. Mann and Sgt. Bailey were j ust naturally a team (either by accide nt or design). Their felld continued, and in add ition a race sprung up betwee n them over the progress made. Bailey never was ve ry coope rative in hi s suggesti ons to Lt. Mann (Bailey was more ad vanced) and one time he nea rl y laughed him se lf into hys te rics afte r look ing over Lt. Mann 's prob le m, g lancing out the window as if viewing the weather prospects, making a few po inted observati ons with a knowing inflecti on, and watching Lt. Mann swallow the bait and madl y try to fo llow the "cl ue" which was entirely fa lse . Sgt. Bailey we nt racing down the hall fair ly shout ing, " Boy, d id I fo ul Mann up! He' ll be there the rest of the winter."

***

82

SIGNALS INTELLIGENCE

In the spring of '43, the going about camp really got tough , with mud about a foot deep every tum one took. By this time the mess hall was open (Christmas dinner was the first meal served in it), and the school shift was the conventional A shift from 0645 to 1245; B shift from 1380 to 1900. The first shift was marched to and from school by T/Sgt Van Cleaf, on detached serv ice from the Air Force somewhere out in Washington State, and the second shift by Sgt. Afflerbach. Every morning there was the problem of finding a route to school that invol ved no sw imming to cross the road in front of the E buildings. Thi s problem and the routes finally chosen confused Van Cleaf who mi ssed the tum , went out the main gate, marched down the road (all of this in total darkness), and arrived at school twenty minutes late. Afflerbach , having the noonday sun to mane uver by, did nothing worse than lose a few people when they slipped off the foot bridge over the mud in front of " E" Bldg. 2. With the coming of spring came a great influx of students and the need for more in structors. As a breakin g- in course, the embryonic professors were first g iven the test of showing their ability with the students in Room 3 1.... No one ever was compl ete master of Room 3 1, because the Schwartz boys, Duke (of latrine map fame) , and Smitty (Lou) were just invincible. Many are the times that their stories have held up before Capt. Guerney's questioning, though everyone knew they had been in Warrenton long before noon that Saturday. Tarpley and Hodge, rather than face Room 3 1, chose to become warrant officers and take part in the invasion of Sicily. During this phase of the campaign, military training was strictly a hit or miss afair supposed to take place during the shift that the men were not in school. What with the large number of men on detail and the "goof off's." very few were out at anyone formation. To counteract the ri se in goofing off, Griffin and Afflerbach cooked up a surprise roll call one morning and found that the absentees just about balanced the presentees. Before the inevitable three days KP, the boys admilled that they had been ( I) in the barracks, (2) in the bushes, (3) in the PX, (4) in "A" Company asleep and unnoticed because of "A" Company's 24-hour schedule. In May 1943, enough buildings became available to discontinue the two shift basis and also to crack the tradition of the thirteen day week. The training program no w was held in the afternoon as the finish of a day and utili zed the Kun zman Memorial Obstacle course, the Nason grape vine hang-on-or-get-wet-course, and Sgt. Neuwoehner's tick-infested "creep and cra wl" course. After the demi se of Ephron , Clark , Latham, Lehr and Schier, Gramann and Chapman, wi th Sgt. Nelson as boss, took over the paper grading department and worked madly trying to make at least thirt y new problems per day in addition to going over hundreds of papers. With 147 sq uare feet o f map behind them, thi s " General Staff' was masterminded by the exci tabl e, flu et-playi ng, Esperantospeaking, Egypti an-born Greek , Lt. Callimahos who s imultaneo us ly prepared problems, practiced Morse Code, worked on in ve ntions, played chess, and listened

TRAINING THE FORCE

83

in o n other sections' conversation. Despite a sharp inte ll ect , the Lt. had a g ullible streak that was ex ploited by his buddies whenever the situation was ripe (as it usually was) .... The rest of the s umme r of 1943 was quie t. Gus Uhl we nt to Canada just long enough to get a service ribbon and then retumed ; Corry set a new beer record in the 130 pound class and added acrobatics pe rformed on the railing around the PX; Palme r 's room o f c ripples were all e ithe r declared fit or sent home to find jobs, thus free ing him for active duty ; and Lamke (Bk s. 13 all present o r accounted FOO-URP!") was quietly bitching away in th c library, abl e assisted by Oscar Briggs. Be rshas and Wi gg in transferred their efforts from the language field, probably to fill the gap le ft when Oppenheimer went to the Southwest Pacific.

*** Besides o ther school equipment . a complete staff of able officers was around at all times. Some made the complete cycle of student, instructor, company executive officer, and back to student. ... " Bi g" Bundy and Lt. "Little" Bund y conspired toge the r with some message center problems that seemed to be always planned for the coldest and most un comfortable day of the month ; and although they ' ve been go ne for nearly two years, there still stands on the main gate road the buggy little shack they immortalized as "THQ"-a monument to their work. Autumn of 1943 brought big thin gs and a lot of new instru ctors. Jun gle Jim Metca lf took on instructorial duties above and beyondthe call of malaria control; Elgin and Braswe ll were sa lvaged from the Room 31 ga ng; and the AtamianMarburger combine was the result o f random se lection . The " quickie" course came bounc ing off the press (really from the pens of Lt. Callimahos and Gerlach) and the first ex perime nt at force feeding was unde r way. Very few chok ings or strangulations were reported. At the first anniversary at Vint Hill approached, plans were made for a celebration of those who had made the initial trek and were still hangi ng around. The dinner was he ld at the Robert E. Lee, one of severa l outposts of Greek c ivili zation in Warrenton , on the evening o f October 5, 1943. Seven teen of the twe nty eligibles enjoyed the spaghetti dinner and then retumcd to Room 3 18 for coffee (and bourbon , with poker). Sgt. Kruege r and Owens, as usual , finan ced the game and finished by paying off Cpl. Bill Hewitt 's wedd ing ex penses. Bill Doering de monstrated a joumalistic nair by writing hi s first ful l le ng th nove l, "Your Life at Vint Hill ," ... a memorial which gave rise to the O ld Bastards Club.

*** By May, the school had dwindled to the point where each instruc tor had hi s own student ; many of the Warrenton Commandos saw fit not to plant a garden ; and

84

SIGNALS INTELLIG ENCE

everyone was tryin g to ge t in a co upl e of ex tra courses that would he lp hold stripes . Came the '" bust" in ASTP, and the n the horde. Every avai lable room (School Section '" B" was even declared bankrupt) and spot was utilized for classes, with "The Dome" Rosskopf hav ing the misfortune to draw the Salt Mines (Suppl y Building) whe re the tar on the roo f melted and spallered freely. He finally got a helpe r in Les Olin. but Olin decided if he could stand tar he could stand anothe r substance and went to OCS .. .. On Jul y 4th , (ah, what a day for fireworks!) the first class of WACs came to School Section "A," and what a change in the EM!' Students no longer had to be rem inded 10 shave, and it is said that the sa le of Kreml quadrupled at the PX . The one sad note in the coming of the WACs, other than the loss of one perfectl y good latrine by the EM, was that , after many long hours of planning and master-minding. Afflerbach was at the las t moment informed that a more stabl e and experienced man would be required to master the co-eds and that Sgt. Ne lson wo uld fit exactl y (as long as his didn ' t find o ut). The WAC occupation can be divided rough ly (or perhaps gentl y) into three periods: ( I) the Nelson or pioneering period; (2) the Ramsey-S nyde r or achievcme nt period; and (3) the Palmer o r decade nt period. During the first of these, the personnel of the gro up exhibited a timidity, seriousness, and lack of humor for which Nelson claimcd credit but whic h probably refl ected the newness of their situation: for although the natives looked upon them with interest and c uriosity, they treated them with fonnality and restraint. During this period of plain living and high thinking, the first personalities to emerge and take definite shape (uh huh!! ) were those of Mary Jane, the heroine. a nd Sue the siren. The classroom held an odor variously interpreted as (a) C hanel Number 7 and Tabu , and (b) G I soap and honest sweat. Toward the e nd of the period a feud wit h the g uards resulted in a marching forma tion that was practically all road guides. During the second period the gro up became more self-assured, both in the ir work and in their soc ial relations. During breaks they mingled with the male students, who began to accept the m, and their fears regarding the apparent ab norma lity of the male behavior were quelled as the sheep's clothing graduall y di sappeared. As the g irl s allacked their work with mo re confidence, Summerhalter and Mo rri s became adept at substituting charm for achievement. McMann showed a cert ain athletic prowess while wers tling during breaks, and sedate M/Sgt. Filipczak was occasionally moved to execute trapeze ant ics on the railing outside the door near the classroom. There is no positi ve proof that Sgt. Palmer's presence ca used the peri od of decadence, but it is unfortunately true that man y of the brighter fl o uri shes OCC UlTed during hi s brief reg ime. The trouble waS that AR's and SSAR's (School Section "A"' Regulati ons) had never e nvisaged the e ntry of females into the school, so that many of the lillie incidents had to pass unregulated. For example. what could be done when Gi nn y Blakemore began to put up he r ha ir during class, or whe n

85

TRAINING THE FORCE

Summerhalter showed up with curlers in he r hair? Both continued working, so not even the crime of loafing could be pinned on the m, And the cigarelle holders nearIya foot long - stri ctl y not GI, but apparently not contrary to the rules, Bell got closer 10 the borde rline when she impaled a marshmallow on a pencil, the n toasted it over a match, but she was ahead of the rest o f the class and had earned the right to an eccentriciry or two, SgL Palme r felt that much of the tro uble was really the fault of SgL Nelson , for everything he (Palme r) did was wither app roved by a chorus of "That 's just what SgL Nelson said", or di sapproved by the accusing re minde r that "SgL Nelson never spoke to us like thaI, " The story had a happy ending, though, for on hi s birthday the g irl s sang "Happy Birthday, SgL Palmer", and have him some cigarelles, candy, and aspirin- the laller to remembe r them by, they said, As the second year staggered to a close, the Old Bastards held a second meeting, again pUlling out a me morial of the occasion .... Although both Bastards and bourbon were scarcer, the party demon strated that the aging me mbers had lost none of the capacity for ce le bration. They wondered if they (or anyone) would be a round for a Third Anniversary. It didn't seem likely, but then as LL Neuwoehner used to say, " We' ll never touch Palau and Halmahera' "

***

CHAPTER V

Cracking the Codes The principal Army cryptanalytic center in World War U was located on the grounds of a former women's junior college at Arlington Hall Station in Arlington, Virginia. The main school building housed the headquarters of the Signal Intelligence Service (later redesignated as the Signal Security Agency); operational personnel worked in two large and unlovely temporary buildings run up in ninety days in 1942 and 1943. By the time the war came to an end, 5,000 civilians and 2,000 military personnel were work ing around the clock to decrypt Japanese military, diplomatic, and other messages. The excerpt from SRH 349, Achievements of the S ignal Security Agency in World War II , g ives some details of their accomplishments. A message reproduced from SRH 280, An Exhibit of the Important Types of Intelli gence Recovered through Reading Japanese Cryptograms, demonstrates that solution of Japanese codes and ciphers could even produce important intelligence dividends on the war in Europe. Other Arnly cryptana lysts were located at Bletchley Park , the headquarters of the Government Code and Cypher School (the British signals intelligence organization); with fie ld units; and at General Douglas MacArthur's Central Bureau, an Allied crypto logic unit originally headquartered in Brisbane, Australia. A short history of Centra l Bureau can be found in the excerpt from A Brief History of the G-2 Section, GHQ, SWPA and Affiliated Units: Introduction to the Intelligence Series (General Headquarters, Far East Command, Military Intelligence Section, General Staff), pp. 66-69.

SIGNALS INTELLIGENCE

88

From SRH 349. Beginning in 1942, the Signal Int elligence Service was repeatedly reorgani:ed and redesignated, finally iJecoming the Signal Security Agency, a designation wh ich oiJscured its major mission.

THE ACHIEVEMENTS OF THE SIGNAL SECURITY AGENCY IN WORLD WAR II

I. BUILDING THE ORGANIZATION The Signal Security Agency (SSA) owed its existence in World War II to the basic fact that valuable information may be derived by intercepting communications and reducing them to intelligible form. This fact has two aspects: efforts must be made to protect our own communications against examination by the enemy, while at the same time steps must be taken to derive as much infonnation as possible from enemy comm unications. The SSA had primary responsibility for both these phases.

*** In spi te of greatly limited funds, the SIS was able in the years prior to World War II to lay securely the foundations upon which wanime expansion of the SSA was built. ... Activity of the SIS consisted of continuatio n of ex isting code producti on projects; establ ishment of a training program by whi ch a small g ro up of expen c ivilian cryptana lysts was produced, and another small group of officers, both Regular and Reserve, were trained in the varied phases of signal inte lli gence; development of intercept facilities to provide material for the cryptanalysts; and, toward the end of the period, the reestablis hmen t, as a practical operation , of so luti on of current diplomatic traffic. In add ition, time was found to do much planning for war and to carryon research and development in cryptographic and cryptanaly ti c techniques. All of this work was done by a very smal l staff: Only seven persons from 1930 to I 936--by the outbreak of the War in Europe there were on ly nineteen. Thereafter, a series of expansions resulted in the following strength on the day of the Pearl Harbor Attack: Category Officers Warrant Officers Enlisted Men Civilians Total

ill Washington 44 0 28 109 181

ill the Field I 0 149 ~ 150

Total 45 0 177 109 331

CRACKING THE CODES

89

Actual operating strength was somewhat less, since 22 of the civilians were still undergoing training and had as yet made no contribution to the work. A comparison of these figure s with the strength of v -J Day ( 14 August (945 ) will show the tremendous proportions reached by the wartime expansions: Category Officers Warrant Officers Enlisted Men Enlisted Wome n Civilians

Total

ill Arlington 661 4 565 957 5661 7848

In the Fie ld 116 II 2139 257 __ 0 2523

Total 777 15 2704 12 14 5661 10,37 1

These figures do not include, of course, the more than 17,000 officers and enlisted personnel engaged in signal intelli gence activities unde r overseas theater commande rs, nor do they give any adequate indication of the tumover of personnel in the Military District of Washington: by the end of the War more than 18,000 numbe red badges had been issued to all categories of pe rsonnel at Arlington Hall (headquarters of the SSA ) alone. To recruit such a staff and to maintain it despite the man y influences which tended to di ss ipate and to lower strength were tasks requiring strenuous effort s.

The rec ruitment prog ram was faced not only by obstacles which also beset othe r wartime age ncies, such as the manpower shortage and the need for speed, but also by many problems peculiar to the SSA, most of the ac ti vities of which had no counterpart outside the Govemment from which to dra w personnel. Many of its operations required persons of the highest intelligence, possessed of rare skill s not easily definable, often demanding, as in the case of the language expe rts, yea rs of stud y to produce proficiency. Messages in more than [text withheld] languages had to be translated, yet in only a few instances were competent linguists eas ily obtainable. This problem was most acute in the case of Japanese, both because a knowledge of Japanese is rare in thi s country and because the volume of mate rial to be translated was so great [text wi thheld] The only recourse was to train pe rsonnel from the ve ry beg inning: ultimate ly, for example, 428 Japanese translators were trained in this way. The same method proved to be the only solution for supplying com petent cryptanalysts: both by the use of training manual s in military cryptography and cryptanalys is and by apprentice training in operating units, the small number of compe tent cryptanalysts available in 1941 was grad uall y expanded, but the suppl y was never eq ual to the demand. Moreover, the SSA was called upon to train large numbers of personnel for ultimate ass ignment to overseas units maintained by thea ter commanders. This training was given not on ly in forma l courses designed to produce officers and men qualified in the various cryptographic and cryptana lytic specialties but also by

90

SIGNALS INTELLIGENCE

participation in the day-to-day activity of operating units. Much was done also to coordinate training of signal intelligence and radio intelligence units being trained elsewhere in order to keep them abreast of the latest technica l developments. In this connection it will be well to digress long enough to point out the trend exhibited during the War of breaking down the centralized authority of the SSA by setting up independent signal intelligence units under theater commanders . Such a change was motivated no doubt by a feeling that it would be necessary to maintain signa l intelligence units close to military operations. Yet the experience of the War showed that modem electrical communications are so speedy that distance is no longer a factor that need be considered. Examples will be cited ... of messages transmitted many thousands of miles, promptly read and translated, and sent back to the proper commander in time for action. Morale problems were particularly acute, for in addition to those encountered by other wartime agencies. involving living conditions, health. fatigue , and the like, the SSA had a number peculiar to itself. Chief among these was the mixed character of the personnel. Officers, enlisted personnel, and civilians worked side by side, and in each of these groups there were both men and women. A small unit composed of personnel from several of these groups might contain people having varying degrees of prestige and receiving different pay and privileges, yet performing exactly the same type of serv ice .... Moreover, there seemed to be no correlation between a catego ry of personnel on the one hand and competence and achievement on the other. Brilliant work was done by individual s in all categories. Had it been possible to operate the SSA entirely with military personnel , or entirely with civilians, some friction might have been avoided, but an SSA made up only of military, or only of civilians, would have lost immeasurably the contributions of the other group. As it was, military personnel had to be stationed with the SSA for both operational and training purposes while having civilian employees made it possible to use the services of many not qualified for military duty. Another factor not particularly conducive to the maintenance of high morale was the necessity of maintaining complete silence concerning every phase of one's work. Moreover, many of the operations involved nothing but drudgery and considerations of security prevented the individual employee in many cases from getting a clear understanding of how hi s or her work contributed to the war efFort. Continuous efforts were made to maintain morale at a high level by bettering conditions of work and furnishing employees with aid in the solution of their personal problems. While the rate of separations of civilian employees (4 per cent a month) seemed high, it was found to be actually lower than that of other comparable wartime agencies in Washington. Indeed, giv ing due consideration to the inherent difficulties, morale in the SSA was really high. Early in the War space in the Munitions Building, where the SIS had been located from its founding, grew so crowded that larger quarters had to be secured elsewhere. At first it was expected that the SIS would be housed in the Pentagon,

CRACKING THE CODES

91

tben nearing completion, but before such a move was effected, plans were laid for moving the SIS to a site of its own. preferably outside Washington. The SIS would thus have room for expansion, be relatively beller protected from the danger of enemy bombing, and could maintain its security with greater ease if not forced to share a bui lding with other War Department agencies. The site ultimately chosen after examination of several possibilities was the premises of Arlington Hall Junior College in Arlington, Virginia. This location was close enough to the Pentagon. was not too far from the Eastern Primary Monitoring Station then being planned at Vint Hill Farms, near Warrenton, and in addition made possible the utilization of the housing facilities in the Washington area for quarters for personnel and dependents. The property was acquired by court action for $650,000, with $40,000 additional for furnishings, and the War Department assumed possession on 14 June 1942. Arlington Hall Station was established as a Class IV installation under the Chief Signal Officer on 25 June 1942. Immediately units of the SIS began to move into the former school building, now the Headquarters Building, Arlington Hall Station. and by 24 August 1942 all of the SIS was at Arlington Hall , the move having consumated without interruption of operations. Construction began almost immediately on projects designed to convert certain of the existing buildings to military purposes, and by September the program of new construction was in full swing. Bes ides many smaller buildings, two very large semi-permanent buildings were erected for operations. The first was completed in less than three months time; the second, erected in winter months, took a little longer, but after May 1943 all operating units were housed in the two operations buildings. Other construct ion provided barracks and mess halls for enlisted personnel. a stat ion dispensary, theater, post exchange, and other service buildings; a motor pool , a fire house and warehouses. The final construction was completed by 1944: while many units have had to work in crowded conditions, adequate space was supplied for all operat ions. One of the important anci llary structures was a cafeteria , completed early in 1943 and enlarged in 1944. which ultimately provided service around the clock .... The Post was made secure by establishment of a guard detachment and the erection of a double steel fence provided with an alann system. Another fence was erected around each operations building entrance to which was limited to authorized personnel. Distinctive badges were issued to all categories of personnel and had to be worn at all times .... Access to the Post was forbidden to visitors except on official business, and efforts were constantl y maintained to indoctrinate all personnel in the necessity of maintaining the physical and operational security of the Agency. The S IS had been prior to the War a field service under the Chief Signal Officer. Though located physically in Washington, it was not a part of the Office of the Chief Signal Officer but was administered at first through its War Plans and

SIGNALS INTELLIGENCE

92

Training Di vision and later through its Operati ons Branch. With the coming of war, however, reorganizations were made by the Chief Signal Officer which result ed in a series of changes of name. The o ld name of Signal Intelligence Service (SIS ). whi ch had ex isted si nce 1930, was aband o ned fo r Si g nal Inte lli ge nce Division (S ID), and in rapid succession, thi s became the Signal Security Branch (SSB), the Signal Security Service (SSS), and , finall y, on I Jul y 1943, the Signal Securit y Agency (SSA), a name whi ch remained unchanged until the cessation of hostilities. After the organization of the Army Communications Service within the Offi ce of the Chief Signal Officer, the SIS-SSA , by whatever name it was called, was always a pan of the Army Communications Service. The need fo r enl isted personne l was met by increas ing the streng th of the Second Signal Service Battalion. This Battali on had been created as a company on I January 1939 to suppl y personne l fo r the various intercept stations then in ex istence. Later it furni shed a convenient military un it to which enli sted personnel employed in the Military District of Washington and elsewhere for other types of signa l inte lli gence ac ti vity co uld be ass ig ned . In November 1942 the post of Command ing O ffice r of the Batt alion was united with that of the Command ing O ffi cer, Signal Securit y Agency: in this way it was poss ible to effect unified control of all SSA enli sted personnel wherever stati oned. The Battalion had detac hments not onl y at Arlington Hall Stati on but all over the wo rld wherever it became desirable to conduct any intercept activ ity. The Battalion did not, of course, inc lude sig nal intelligence personnel directl y under the control of theater comm anders. While the Battalion embodied man y unorthodox features-it s maximum strength, for examp le, surpassed 5,000 officers and men and at the end of the War it was commanded by a Bri gadier General-it effecti vely solved the problem of how to admin iste r the am az ing ly compl ex ac ti viti es performed by e nli sted personne l working for the SSA.

*** It has already been noted that the SSA was adm inistratively and fun ctionally a pan of the Signal Corps, but G-2 exercised staff supervision and control. While channe ls we re erected whereby G-2 could exercise thi s control without at every step go ing through the Office of the Chief Sig nal Offi cer, neverthe less the SSA was primaril y a Signal Corps age ncy, its personnel we re Signal Corps empl oyees, and for purposes of suppl y it re lied on Signal Corps fac iliti es. Since it was increasing ly felt by G-2 that the SSA was the most important so urce o f int e lli ge nce, eve n c lo se r co ntro l was required and th e re fo re, o n 10 December 1944, there came a change. The SSA was removed from the Si gnal Corps for operationa l contro l, which was now ass umed by G-2 , but admini strati ve cont rol was still retained by the Signal Corps. Thi s cleavage of contro l was by no means c1earcut and sharpl y defined: the line of d iv ision was not straight , since the

CRACKING THE CODES

93

organization was pragmatic rather than theoretical, but in the main the differentiation thereafter was operational control exercised by G-2, administrative control maintained by the Signal Corps. Divided control. such as thi s was, however, proved far from sati sfactory in practice and was ended on 15 September 1945 by transfer of administrative control also to G-2, with the change of name from Signal Security Agency to Army Security Agency (AS A). This had the effect of associating the organization more closely with G-2, the user of one of the two chief products, but it will necessitate that in the future liaison be constantly maintained with the Signal Corps so that not only an adequate supply of trained personnel may be available for the communications side of its activity but also the increasingly closer relationship between signal intelligence and signal sec urity may be maintained by the closest cooperation. Moreover, it also resu lted in once more consolidating all responsibility for signal intelligence and signal security in a single organization, since the units formerly under the control of theater commanders were now made a part of the ASA. As will be abundantly clear from spec ific references in the two following c hapters, the SSA had the incalculable advantage of collaboration with the corresponding units of the United States Navy (OP-20-G and OP-20-K). Liaison with the Navy had long been in progress before the War but throughout the conflict it constantly increased in both the cryptographic and cryptanalytic field s. For some years before the' War the Army and the Navy had been collaborating [text withheld] but the Office of Naval Communications, being pressed for personnel and facilities needed by unit s at work on enemy traffic [text withheld] asked the SIS to take over more and more work .... Early in the next year the first SIS mi ss ion was sent to England to establi sh the basis for liaison and in the summer the SIS and GCCS first exchanged permanent liai son officers, a relationship ever since maintained . Special missions have, however, also been sent and received from time to time. Intercommunications by radio, cable, and mail, have constantly maintained. Frequent agreements have been made to avoid unnecessary duplication of effort: the chief of these was reac hed in 1943 whereby the British assumed primary responsibility for signal intelligence operations for the War in Europe, the SSA , for those for the War in the Pacific, though neither Agency abandoned work in the field of the other's responsibility. This exchange of information has been broadest in cryptanal yti c activity: considerations of security ha ve limited cooperation in cryptographic compilation and development to work on systems used in combined British and American operations .... [text withheld] The SSA provided trained personnel for and collaborated with U.S. Army Signal Intellige nce Services in all theaters: Mediterrean Theater of Operations, European Theater of Operations , Southwest Pacific Area, China-Burma-India Theater, Central Pacific Area, etc. Collaboration between the SSA and the Central Bureau, Brisbane (CBB), began with the founding of the latter organization by joint action of the Royal Australian Army and the United States Army in the spring

94

SIGNALS INTELLIGENCE

of 1942. As will be later seen, this cooperati ve effort was maintained by constanl intercommunicati on, particularl y in the case of the Japanese Anny cryptanal yti c problem.

*** II. THE PRODUCTION OF INFORMATION Modem intelligence servi ces are able to derive info nnatio n usefu l for military purposes from man y so urces but th e m ost fruitful and mos t au th e nti c is th e enemy's message traffic and the communication sys tem over which it is tran smit-

ted. Indeed, thou gh messages differ in va lue, experi e nce has proved that the re are no messages, no matter how insignificant in content , whic h have potentially no inte lligence value. Therefore, the cryp tanal yti c attack had to be made not only upon the purely military traffic transmitted by e ne my forces but also upo n diplomatic, commercial, and private messages as well. Even plain-text messages could not be neg lected, but the largest part of the intercepted traffic was, of course, in c ryptographic fom1 and required cryptanal yti c treatment before it could be read.

While some of the techniques used by the SSA were no t strictly cryptanalytic in character, e .g . traffic analys is, secret ink solution , and exploitation of te lep ho nic communi cations, the production of info nnation involved , in general, the following steps: a. Interception of traffi c in large volume; b. Traffic analysis of intercepted messages; c . Solution of the cry ptog raphic systems used; d. Decryptographing of messages sent in so lved or partially solved syste m s; e. Translation of such texts as were in foreign lang uages, and f. Publication of the tex ts in a fom1 useful to the Milita ry Inte lli gence Service.

*** A. INTERCEPTION No cryptana lyti c attack upon the communications of a foreign govern ment can hope to be successful unless an adequate suppl y of inte rcepted material is available for stud y; nor can proper measures for sa feguarding our own communications be

taken without constant monitoring of the trafric sent out by American stations. It therefore became necessary to establish facilities for the interception of radio traffic in large vo lume. Prior to the war the re were seven fi xed intercept stations located as follows:

CRACKING THE CODES

95

No. I Fort Hancock , New Jersey No.2 Presidio of San Francisco, California No.3 Fort Sam Houston , Texas No.4 Coroza l, Panama Canal Zone No.5 Fort Shafter, Territory of Hawaii No.6 Fort McKinley, Philippine Islands No.7 Fort Hunt, Virginia Constant efforts during the War ex panded these facilitie s greatly. In the end there we re eleven fixed stations, many of which were far larger than any operating in 194 1. These eleven , which were found suffi cient to supply the necessary volume of traffic, were distributed as follow s: No. I Vint Hill Farms, Warrenton,Virg inia No. 2Two Rock Ranch , Petaluma, California No.3 Indian Creek Station, Miami Beach, Florida No. 4 [text withheld] No.5 Fort Shafter, Territory of Hawaii No.6 Amchitka, Aleutian Islands No.7 Fairbanks, Alaska No.8 [text withheld] No.9 Bellmore, Long Island No. 10 Tarzana, California No. II Guam The three largest station s (at Vint Hill , Two Rock , and Fort Shafter) we re equipped with elaborate arrays of high-directivity antennas for all-round coverage. These stations had been located so as to make easy the electrical forwarding of the intercepted traffic to Arlington Hall and the largest portion possible of intercept missions was assigned to them. The supplementary stations, particularl y those at !text withheld] Amchitka, Fairbanks, and [text wit hheld] were located so as to intercept signals which could not be copied at the larger stations, and in general they had antenna systems beamed at specific targets of sectors. The station s at Bellmore and Tarzana were assigned the task of monitoring United States traffi c for security purposes . Considerable assistance was rendered to the intercept facilities of the SSA, particularl y in the period be fo re the SSA's own intercept facilities were fully developed, by radio intelligence companies stationed on the West Coast and in the Pacific Area. Among the new items of equipment developed during the War for use at intercept stations were "multicouplers,"' which allow the signal from one antenna to be coupled to several receivers; a "Hell schreiber Facsimile Recorder," for copying signals of thi s Gennan system; and a "Time Delay Dev ice," which accomplished a delay of from three to ten seconds between the time a signal is received and the time it is necessary to copy it, making it possible to start a recorder to take down the entire transmiss ion for later transcription.

96

SIGNALS INTELLIGENCE

Intercept activity was coordinated and contro lled by staff units at Arlington Hall which supplied the stations with technical advice. Speedy transmission from the intercept stations was effected chiefly by special teletype lines, which came more and more to take precedence over other means such as cable and air mail. From four teletype lines in operation on 7 December 194 1, the number of such lines grew until on V-J Day there were forty-six . The amount of money paid for monthly rental of land-line teletype facilities alone reached in August 1945 the large sum of $58,9 18.02 but this figure does not include the cost of radio-teletype faci lities, paid for by the ArnlY Communications Service, for which no data are ava ilable to the SSA. Average monthly volume also constantly grew: December 1943 279,034 messages February 194346,865 messages July 194538 1,590 messages August 1945289,802 messages Arrangements were also made for obtaining traffic from radio intelligence units operating in theaters of wa r; from the Navy; from the several offices of the Chief Cable Censor; [text withheld] B. TRAFFIC ANALYSIS Traffic analysis, a procedure which first arose from attempts to reconstruct enemy communications networks and their characteristics with the aim of improving intercept facilities, became highly useful also for two other purposes: ( I) Through study of the external features of the messages as distinct from the text itself, together with direction finding , by which it is poss ible to locate the site of unknown radio stations, traffic analysts were able to provide cryptanalysts with much useful information not otherwise obtainable; (2) Statistical study of the fluctu~ttion s

in the volume of traffic pass ing in each circuit , and inferences drawn

therefrom , became an important source of military intelligence. Traffic analysis can be carried on, of co urse, independently of successful cryptanalysis: use ful infonnation can be derived by traffic ana lysis even before a message is readable, but when the two techniques are combined, each is aided by the other. While traffic analysis had been used to a limited extent in World War I. the British were the first to develop the science extensively in World War II. The beginnings of traffic analysis in the SIS date from April 1942. A mission was sent to England to gat her infonnation and upon its return it was possible to set up traffic analysis as an integral part of the SIS. As a result of this mission, efforts of the SSA in traffic analysis were to be concentrated on traffic in the Pacific theater, leaving to GCCS the primary responsibility for that in the European, a logical arrangement arising from geographical considerations. The initial problem in traffic analysis for the SSA was the solution of the code numbers used to indicate message center place names occurring in Japanese mili-

CRACKING THE CODES

97

tary messages, and the first success was achieved in September 1942. By the fol lowing June nearly all of the twelve main systems had been reconstructed, permitting accurate location and mapping of radio stations and circuits. Four distinct major military networks were identified , those used by the Imperial GHQ in Tokyo, the Southern Field Force, the Wate r Transport organization. and the Army Air Force. The adequacy of the teChniques used was proved when, on I April 1944, the Japanese introduced a completely new place-name code which was almost wholly solved within a month , about half of the names being identified within 48 hours. Technical assistance given the intercept stations was responsible, at least in part, for the rapid increase in volume of Japanese military intercepts. Contributions of traffic analysis to military intelligence began with the location of military message centers attached to unit headquarters. The ide ntification of nets and unit organization revealed troop locations and chains of command, thus giving highly important information concerning the Japanese Order of Battle. Traffic flow analysis-the study of variations in traffic volume and pattern s of station activity-gave indications of impending enemy activity. Convoys were detected and fol lowed merely by studying the changes in the pattern of communications in the Water Transport code between various seaports along the route. Analogous studies of patterns in Air Force and Army Administrative codes led to detection of aircraft and troop movements. Analysis of communications between field units and their home depots indicated the location of almost all divisions south of Manchuria. The movement of a unit in the field could be detected from its home depot traffi c, or through readdressed messages, or by messages addressed to the unit code name in li eu of a specific message-center location. Finally in sight into the general content of diverse classes of messages, gained through traffic analysis, aided in establishing priorities in handling the thousands of messages received daily. Though publi cat ion of daily and weekly traffic analysis bulletins began in September 1942, the greatest emphasis on the intelligence aspect of traffic analysis came in late 1944 and 1945. Where deceptive measures are not employed. traffic analysis can yie ld a very large amount of fairly reliable inte lli gence; where deception is effecti vely practiced, deductions from traffic analysis must be used with ex treme care. Since, however, there was c urrently no evidence that the Japanese engaged in deceptive communication measures (a fact which was confi rmed after V-J Day). valuable results from traffic analysis were frequently obtained . C. CRYPTANALYSIS Cryptanalytic procedures are never stereotyped and permit no easy description . . . . Ilext wit hheldl Too much emphasis cannot be placed upon the need for cryptanalytic continuity . ... Itext withheld] It is characteristic of most cryptographers that they tend to introduce new cryptographic features and elements in a conservative manner. Thus. a new system will frequently not represent a radical departure

98

S IG NALS INTELLIGENCE

from its predecessor but will , rathe r, be merely a refine ment and improveme nt o f what went before. S ince thi s is likely to be so, the cryptanalyst who can start his stud y of a new syste m with a good acq uaintance wi th the c ryptographic charac ter of the one just made obsole te, and o f othe rs used conc urre ntl y by the same governme nt, will be at a great advantage. The basic fac tor unde rl y ing a ll successful cryptanal ysis, howe ve r, is constant watchfulness for significant details which may provide an entry: one or another of the ope rati ons mentioned above may re veal the natu re of the syste m but the surest method is stati stical analys is. In Wo rld War I and for many yea rs the rea ft e r, suc h stati stica l analys is was de pende nt upon hand methods alone . The cryptanalys t or hi s cle ri cal ass istant s we re fo rced to make frequency counts or othe r stati stical tabul ations o f the unit s or text (letters or di gits), ta ken either singly or in groups, by hand. Not onl y did the process involve an imme nse amo unt o f drudgery but it was also easily subject to serio us e rro r. The eye had to keep the place in a meaningless text while the hand made the necessary indications on the c harts. Naturally, the work , to be de penda ble, had to be pai nstaki ng ly acc urate: the result was that it was time-consuming. Even unde r the conditions o f World Wa r I, when the volume of inte rcepted traffic was relati ve ly low, thi s was already a problem of most serious pro portions. As a consequence, ste ps were taken during the pe ri od of peace to find adequate solutions to the problem . The most signifi cant tre nd in c ryptanalytic research and development during World War II has been, in fac t, the extent to which machine ry has been used to speed up hand methods and al so to pe rfo nn operations whic h, because of their magnitude, could never have been attempted without such apparatus. A ve ry large measure of the success of the SSA in c ryptanal ysis must be attributed directl y to this fac t. The enormous increase in the volume of intercepted tra ffi c wo uld alone ha ve made hand me thods who lly inadequate to acco mpli sh the task se t us in th is War. Fortunately, none of the enemy nati ons seems to have rea lized the possibility of developing such machinery a nd the necessity of protecting their cryptographic systems aga inst attack by suc h means, or if they did, as was true in the case of the Ge rnlans. the rea lization was not too clear and insuffi cie nt emphasis was placed upo n the developme nt of mechanical and e lectrical analytical eq uipme nt. The SSA. on the othe r hand. has had to keep constantl y in mind, while deve loping c ryptog raphic syste ms for our own use ... , the possibility that other nati ons might also make. during th e war or later. simi lar adva nces in machine

crypt analytic techn iq ues and to prepare against that contingency. Thus, any new deve lopme nt in crypt analytic techniques has the immediate effect of causing a converse develo pme nt in cryptographic techniq ues. and vice versa. Thi s is the fund amcntal reason why all researc h and deve lopment in both fi e lds mu st be carried on within a single organization. [text withheld] Several types of mac hine ry ha ve been used. The fi rst of these is standard ta bul ating mac hinery, incl ud ing mac hines available on the open market as well as machine ry of the same gene ral type modi fied or

99

CRACKING THE CODES

developed expressly for the SSA. The method involves the recording of data on a card in which holes are punc hed by means of a key-punch machine and the process ing of decks of such cards by a numbe r of other machines. [text withheld\ While few, if any, cryptanalytic units have failed to make extensive use of these machines, by far the greatest employment of the m has been in the solution of the Japanese Army systems. Indeed , the solution and processi ng of any significant quantity of the hundreds of thousands of messages in those systems would have been impossible without these machines. An indication of the growth of the use of tabulating machines of this type by the SSA in the fact that whereas only 13 machines and 21 operators were at work at the outbreak of the War. 407 machines had been installed at the peak in April-May 1945, involving a total of 1275 persons as operators and su pe rvisors. The monthl y rental for these machines reached a peak in June 1945 of $60,982. In addition to standard and spec iali zed tabulating machinery, another important category of machines was that generally referred to as Rapid Analytical Machinery (RAM). Machines of thi s category usually employ vacuum tubes, relays, electronic circuits. and photoelectrica l principles. A number of different types, designed within the SSA for specific operations, were developed and constructed by the SSA it se lf or in cooperation with severa l contractors, and set up at Arlington Hall Station. One cryptanalytic machine, costing almost a million dollars, was basicall y homo logous to an automatic telephone exchange capable of serving a city of about 18,000 subsc ribers, and the SSA had two suc h mac hines. These two machines were capable of pe rforming operations which. if done by hand met hod s, would have required over 200,000 people. A second cryptanalytic machine, specifically designed to perfOnll a certain type of test by means of e lectrical re lays, served as the eq uivalent of 6,000 cryptanalysts; and an improved machine of the same general nature, but using vacuum tubes and electronic principles rather than relays. is now almost completed. It is expected to operate at least 500 times faster than the relay type and can be estimated to be equivalent to having 3.000,000 people at work. In the deve lopme nt and construction of these highly spec iali zed cryptanal ytic machines the SSA expended severa l million dollars, but it could hardl y have operated without them. D. SOLUTIONS Cryptographi c systems have grown in compl ex ity very g reatly si nce World War I. and in a brief report such as this it is impossib le to give more than an inkling of the length to which many foreign government s have gone in improving their methods. Larger and more sc ientifically constructed codes were introduced ; complex radio procedures and supere ncipherme nt syste ms were added; and intricate cipher machines were developed. By the year 1939. when the outbreak of war in Europe caused the first substantial expansion of the SIS since 1930, cryptanalyt -

100

SIGNALS INTELLIGENCE

ical attack was being centered in accordance with directives from G-2 [text withheld] ... all systems thus far studied were entirely diplomatic in character; indeed, little traffic of other categories was then being intercepted. Work proceeded along these lines, effecting so lutions which will be discussed in greater detail a little later, but so far as cryptanalysis is concemed, this was the situati on existing on the day of the Pearl Harbor attack. [text withheld] (I) Japanese Diplomatic and Military Attache Traffic. The first diplomatic systems to receive attention in the SIS were the Japanese, and from 1933 to 1941 eleven such systems were studied and solved. For the most part they were not greatly advanced in cryptography from those so lved before 1930, but in two machine ciphers, the earlier of which had appeared by 1932, the Japanese demonstrated that they had read with profit The American Black Chamber, in which Herbert O. Yardley, formerly an officer of the Military Intelligence Division, indiscreetly revealed to the world American successes in solving foreign cryptographic systems. The machine ciphers presented cryptanalytic problems of greater difficulty, involving not on ly the reconstruction of a complicated machine but also thereafter the day-to-day recovery of a great number of keys. The second of these machines, introduced in 1938, was much more complex than the first and required almost two years of concentrated study to solve: its solution by the SSA; unassisted by any other cryptanalytic organization, represented an achievement of first magnitude and importance. It is now known that the German organ izat ions attempted the feat and failed, [text withheld] The remarkable feature of this solution was that a machine capable of deciphering the Japanese messages was reconstructed wholly by analysis: the SSA has never seen one of the Japanese machines. The importance of this solution can hardly be over-est imated. General Marshall, in his now famous first letter to Governor Dewey dated 25 September 1944, and disclosed in the Joint Congressiona l Hearing in the Pearl Harbor disaster, stated that "our main basic of information regarding Hitler's intenti ons in Europe is obtained from Baron Oshima's messages from Berlin reporting his interviews with Hitler and other officials to the Japanese Government." ... In addition to this machine, which remained in current use until the end of the War because knowledge of its solution was fortunately kept secret during the War, the Japanese a lso began to use for diplomatic purposes a variety of other highsecurity systems based on intricate cryptographic principles. [text withheld[ The result was that the best efforts of a large staff of the ablest experts working continuously on this problem were necessary to solve the diplomatic and military attache systems. Solution was, even so, effected in nearly all cases, providing a very large volume of translated messages giving significant infornlation as to the intentions of the Japanese, conditions in the Far East, and also of conditions in Europe. Indeed, it has been said that the Japanese military attaches were the best secret agents of the United Nations on conditions inside occupied Europe.

CRACKIN G TH E CODES

10 1

The value of the intercepted Japanese diplomati c and military att ache tra ffi c fo r intelli gence purposes can best be illustrated by re presentati ve translations, but of the many thousands of messages which appeared in the SSA Bulletin, onl y a few can be chosen for inclu sion in a brief report such as thi s .... a. Berl in 10 Tokyo, ... translated 4 December 1943, sent in the [tex t withheld] (military att ache system) . ... T he full tex t, too long fo r reproduction here, is a report of a visit made in the fa ll of 1943 by a subord inate o f Baron Oshima to the German wes tern fo rtificati ons. The military information contained in this message was of incalcul able adva ntage to the planning of the invasion of France.... b. Berlin (Oshima) 10 Tokyo, ... translated 12 August 1944, sent in the [text withheld] (diplomatic) system. T his message ... has been descri bed by offi cers in MIS as "worth all the expe nses of maintai ning the SSA ." The text describes conversations with the head o f the Todt organi zation. Albert Speer, in which the latte r re vealed to the Japanese, and, thus to us, highl y important info rmation concerning the production o f munitions in Germany.... c. Hano i to Tokyo, ... translated I Fe bruary 1945. sent in the [text withheld] (d ipl omati c) syste m. Thi s message is important because it reveals that the Japanese were interested in obtain ing uranium . . ..

d. Moscow (Sato) to Tokyo, ... translated 30 Jul y 1945, sent in the [text withheld] (diplomatic) syste ms. T his three-part message, the translation o f which was avai la ble to Pres ident Trum an d uri ng the Potsaam Conference, revea ls the acti vity of SalO, Japanese Ambassador to Moscow. at the ti me of the confere nce ... .It will have been noted that while two of these fo ur messages we re translated within two days afte r they were trans mitted, anothe r took about te n days and the fourth nearl y a month for translation. De lays of thi s kind may be attributed to a nu mbe r of faclOrs. [text withheld] Secondl y, a sharp rise in the volume o f inte rcepts may c reate a backlog of unprocessed messages which may take seve ral days to e liminate. Thirdl y, the re is the question of translation: in the case of Japanese texts, for example, in spite o f the g reat effort s made to train competent Japanese translators the re never were enough of these to keep up at all times with the producti on o f the c ryptanalysts. To prevent hi ghl y impo rt ant messages conta inin g informati on demanding immed iate ac tion by MIS from being laid aside until the infornlation was too late to be useful , a po licy was adopted o f scanning the messages as they became readable in orde r to sort the m according to the deg ree o f urgency. In spite of these diffic ulti es, however. it frequentl y happe ned that messages we re intercepted, decoded, translated, and pl aced in the hands of MI S before their addressees might be presumed to have read them. A conspicuous instance of this kind was the famous message by whi ch the Japanese transmitted through the Swiss Governme nt their inte nti on to acce pt the surre nde r tenns. The fact that the Japanese had accepted the Allied te rm s was known in the MI S several hours be fo re the Sw iss Ministe r was able to give the message to the St ate Departme nt. (2) Japanese Anny and Air Force Traffic. The foll owing paragraph is quoted

102

SIGNALS INTELLIGENCE

from a communique which appeared in The New York Times on 4 September 1943: Wewak : Our strongly escorted medium bombers attacked an enemy convoy of five cargo ships and two destroyers which arrived during the night with reinforcements and supplies for the enemy garrison. Coming in at masthead height, our bombers scored direct hits with I,OOO-pound bombs on three freight transports, each of 7,000 tons, sinking them . In addition, one of the escorting warships and a I,OOO-ton cargo ship sustained direct hits and were left ablaze. Numerous small harbor craft were destroyed by strafing. Intense anti-aircraft barrages were encountered and barrage balloons from ship and shore were employed in an endeavor to halt our low-level attacks. Thirty-five fighters flown in from the rear bases to protect the convoy were intercepted in the air. Twelve of these were downed with eight others probably destroyed and five damaged. Three of our bombers and one fighter were lost. Note that in the entire communique there is no explanation of the method by which the American commander in the New Guinea area learned of the presence of the convoy at Wewak: the impression is given, as indeed was intended, that the good fortune of the bombing mission in finding the convoy solely was the result of chance. Yet this was not so. A message had been intercepted and read by the SSA (on 20 August, nearly two weeks ahead of time) which foretold to the Japanese at Wewak the arrival there of the the convoy on the first or second: .... The message in its English foml was forwarded speedily to MIS and thence by radio to the proper commander for his use. This instance is, of course, only one of many which could be adduced to show how in an age of radio communications the necessity of forwarding an intercepted text thousands of miles to a control agency for decipherment and its return when made readable causes little more delay than would have taken place had the cryptanalysts been at work at the points of interception. In order to prevent the Japanese commander at Wewak from suspecting the truth, precautions were taken , in accordance with rigid regulations, to provide an

additional source of the infonnation which he would naturally suppose to be the only one, namely, reconnaissance planes were sent over Wewak. On many occasions, in facl, American commanders were in possession of valuable information

provided by the SSA which they could not use because to do so would have run the risk of revealing to the Japanese the fact that their secret communications were being read by us. The ability to continue reading the traffic as a whole was often a military objective of greater importance than that involved in the successful completion of a specific mission. The bombing of the convoy at Wewak, as just described, is a good example of the effect of the translation of an isolated message--only the third part of a threepart message had been translated in time-but many messages, which individually are less striking. when taken together and coordinated by MIS, pennit the accomplishment of even more spectacular results. The following paragraph is taken from a memorandum prepared by an officer in MIS (27 March 1945):

CRACKLNG THE CODES

103

Use of Ultra (= SSA ) Information for Attack on Japanese Troop Convoy I. Infonnation received. ULTRA traffic on and shortly after 2 April 1944 revealed Japanese plans to send a large convoy, des ignated as the " TAKE" Convoy, to Halmahera and New Guinea. The convoy, consisting of nine merchant vessels and about twelve escorts, sa iled from Shanghai for the south in the latte r part of April, carrying 12,874 troops of the 32nd Division, about 8,170 troops of the 35th Division, with equipment and other military supplies. Messages furnished the identity of the ships and full details about the troops and cargo loaded on each ship. Traffic analysis disclosed the approximate dale the convoy was scheduled to leave Shanghai for Manila and provided current information on the convoy's approximate position on its trip from Shanghai to Manila. Before the departure from Manila on I May, messages revealed the following infornlation: a. scheduled noon positions for each day from 2 May to 9 May; b. an outline of an alternate route to be followed only on receipt of spec ial instructions; c. a plan to divide the convoy into two groups on 7 May to a point N of Halmahera, one part (presumably the 35th Divi sion ships) scheduled to go on to Manckwari. and the other part presumably (32nd Division ships)scheduled to go on to Wasile on Halmahera. 2. Action taken. Infornlation on the composition. loadings. movements of the TAKE Convoy was forwarded to the appropriate field commands, as it became available. 3. Operational results. On 26 April the convoy was attacked by submarine at a point 30 m. W of Laoag (NW Luzon) and one ship was sunk. On 6 May the convoy was again attacked by submarine 100 m. NW of Menado and three additional ships were sunk. About 4,000 troops together with ordnance and other supplies were lost as a result of those sinkings, including the Commanding Officer and 2,700 troops (substantially all) of the 220th Infantry Regiment of the 35th Division. Two Japanese divisions, both critically needed as reinforcements, were thus decimated and their effectiveness seriously reduced. Both divisions have si nce been met in combat. The above information concerning operational results also was received from Ultra sources. 4. Ultra material used. Preparation of the intelligence dispatched to the field commands required the examination and integration

or a large number of sep-

arate and frequenctly fragmentary messagcs and traffic analysis. The fact that the two examples already chosen were both concerned with the sinking of convoys should not be allowed to give the impression that this was the only phase in which il1lelligence was derived from Japanese Army messages. To quote again from the report already cited as prepared by an MIS officcr: A 28 May ( 1944) message, available I June. mel1lioned supplies needed by the 18th (Japanese) Anny (controlling operations in eastern New Guinea) which must a rrive at Wewak by the end of June in order to be of use in "the attack on Aitape."

104

SIGNALS INTELLIGENCE

In a 24 June message, available shortly thereafter, the Southern Army stated that the 18th Army would attack Aitape. Various other fragmentary messages, all showing that an attack on Aitape was planned, were also received. On 25 June there became available a 20 June message from the 18th Army reporting that it was planning an all-out allack against the U.S. Aitape perimeter. to begin about 10 July and giving the detailed dispositions of each division under the command of the Army, plus the planned operations of each division in the attack. Total strength of the forces involved was stated in the message to be about 20,000 .... All of (this) information was made available to the Commander-in-Chief, Southwest Pacific Area, before the date of the planned attack .... The Japanese allack was made on schedule and was completely defeated with heavy losses to the Japanese. The examples just given show how the product of the Japanese Army cryptanalytic projects was of the greatest value in military operations in the Pacific area. We must now tum to the steps by which the product was made possible. Before the War attempts by the SIS to solve the secret communications systems of the Japanese Army had, for lack of sufficient traffic and cryptanalytic personnel , been fruitless. [text withheld] Immediately after the attack on Pearl Harbor more SIS personnel were assigned to the Japanese Army problems, but lillie could be done except to sort and file the traffic. Thus, cryptanalytic continuity was broken for a time. [text withheld] Cipher solution may be compared roughly to the operation of demolishing a building by undermining its substructure and causing the building to collapse suddenly; code solution is more like a mining operation, the ore is taken out of the mine bit by bit. By the autumn of 1942, however, traffic sources had been established and more personnel had been obtained, so that considerable attention could be given to these problems. [text withhe ld] This historical method, though seemingly a long way round , was really the only recourse and it was fully justified by the results. By January 1943 progress had been made beyond the period of [text withheld] and at the same time encouraging results were being achieved in the recovery of code groups for addresses, which had previously seemed so hopeless a problem. [text withheld] In Apri l 1943 the first break into a current system (that was used by the Japanese Army Water Transport organization was made possi ble [text withheld]. In this instance the entering wedge was simultaneously discovered (almost in the very same hour) by personnel of the SSA at Arlington Hall and of the CB B in Australia. It was most successfully exploited as the result of the constant collaboration maintained by daily intercommunications between the two organizations. The result of this initial solution was a reorganization of facilities within the SSA leading to considerable expansion in personnel and services devoted to the Japanese Army problem. Ultimately, the number at work on these probl em s exceeded that of personnel at work on all other cryptanalytic problems combined, and yet there never were enough trained workers to carry out fully all phases which needed to be done; the man-power situation was such at the time that the

CRACKfNG THE CODES

105

demand for personnel was always g reater than the supply, a nd even w he n the needs were temporarily fill ed the re always was some delay be tween the time new personnel reported for dut y and the time they were ready to participate effective ly in actual operations, because preliminary training was usually essential. By earl y June 1943, two months after the initial break into the system , translation of Wate r Transport messages were being forwarded to MIS . Thereafter, until almost the e nd of the War, the Water Transport system provided not onl y a broad picture of the Japanese AlTIl y shipping organization and activities but also, from time to time, infolTIlation regarding specific operational movements of wh ich the logistics proble ms were disc ussed in Wate r Transport messages. To quote once more from General Ma rshall 's lette r to GovelTIor Dewey: "Operations in the Pac ifi c are largely guided by the infolTIlation we obtained of Japanese deployments. We know their stre ngth in various garrisons, the rations and other stores continuing available to the m, and what is of vast importance, we c heck their flee t move ment s and the moveme nts of their convoys. The heavy losses reported from time to time which they sustain by reason of our submarine ac tion largely results from the fact that we know the sailing dates and the routes of their convoys and can notify our submarines to lie in wait at the proper point. The current raids by Admiral Halsey's carrier forces on Japanese shipping in Manila Bay and elsewhere were largely based on timing of the known movements of Japanese convoys, two of which we re caught , as anticipated, in hi s destructive attacks." Although undoubted ly General Marshall included in this tribute to the signal intelligence serv ices the ve ry outstanding contribution from the signal intelligence service of the Navy, for fl eet move ments are me ntioned, the bulk, however, of the information of the type cited came from the SSA. Solution of the main Ground system was effected by a series of discoveries that began in May 1943 and c ulminated in decipherme nt of messages in September of that yea r. Al so, by the summer of 1943, so lutio n of the address syste m had reached a po int where the addresses of c urre nt intercepts we re readab le. Reconstruction of the address code books, w hi ch had hitherto been carried on excl usively by a Briti sh organ ization in India, known as the Wire less Experimental Center (WEC), was no w undertaken at Arlington Hall. The order of battle intelli gence derived from daily li sts of the addresses of Japanese AITI1Y units was a useful adjunct to the text of the messages and even supplied knowledge of military operations in periods when the messages themselves we re not readable. The reafter, despite repeated c hanges and innovations made by the Japanese in their c ryptographic systems, so lution wa, continuous, though on occasion it was temporarily delayed. The knowledge of Japanese cryptograp hi c practices and previous solution of the basic code book permitted the reading of periods which , from a cryptanalytic point of view, were as difficult as the systems encounte red at the time of Pearl Harbor attack, systems whic h had then been considered hopeless of success.

SIGNALS INTELLIGENC E

106

Two major technical problems which had to be solved in 1944 were the introduction by the Japanese on I August of a new cryptographic practice which disg ui sed the system indicators and a radical shortening of the life of one of the keying elements used in the Administrative syste ms. The uncovering of the disguise in the case of each system indicator was, of course, a prerequisite to the subdivi sion of intercepted messages into their respecti ve systems pre paratory to any other steps towa rd so lving or reading the messages. Though the introduction of thi s feature might have been a major cryptanal ytic di saste r, it fortunately turned out to be on ly a nuisance, because it was at first effected by an insecure method, and speedy so lution was the refore poss ible. Subsequently the Ja panese mod ified the method to the point where, had it been used initiall y, so lution would have been almost impossible. With continuit y of so lution , however, aided by information from cryptographic instruction messages and captured materials. thi s handicap was overcome. The problem presented by shorte r inte rvals (five days instead of three weeks) between changes of keys, was essentially one of carrying on analys is of one of the steps of encipherment with only about a fourth of the traffic previously available. was eventually so lved by a combination of methods, namely, very careful correlation of every piece of inte rcepted traffi c, the use of tabul ating mac hinery, which enabled a tremendously large numbe r of ope rations to be made in a short time, and the use o f photoelectric equipment to exploit phenomena resulting from messages with identical or nearly identical tex t but cryptographed with different keys. Once the Japa nese began to suffer military reverses, their cryptographic mate ri als were frequently captured and these soon came to playa vital role in solution. Especially important was the capture of the basic code books . or of the key books which the Japanese could not easily re place with new editions and which the refore were occasionally continued in effect for some time after their capture. The complete reconstruction of such code books and key books would ha ve been a long and painstaking task which would ha ve resulted in delays in production of intelligence and possibly also in some diminution o f its re liabilit y. Yet the continued capture of such material s was not an unmi xed blessing, since whenever a capture was known or even suspected. the Japanese natura ll y changed as many of their other cryptograph ic material s as possible. Thanks to the care wit h which infonnation obtai ned from the translated messages was used by fie ld commanders, the Japanese seem never to ha ve suspected the possibility o f c ryptanal yt ic compromi se but , as they began to realize that Allied Forces were able to anticipate their plans, they att ributed our success to espi onage activities. The following extracts are taken from a message, inte rcepted and read by the SSA. whi ch was sent from Pinrang to Piru ( RIYA Communicat ions Officer) on 18 December 1944 ...

***

CRACKING THE CODES

107

" . . . . there are s ubs tantial indications that the enemy has unde rstood our important plans in the BUllna and Philippine areas. Therefore, we are inclined to be somewhat doubtful about the codes now in use (and) each unit commande r must multiply his alertness toward counter-espionage .... This is an order. Furthellllore, yo u should exert yo ur best efforts towards overcoming the deficiencies in the counter-espionage set-up, and towards perfecting it. If yo u fail to do thi s, troubles will arise, and yo u must take reso lute action in facing them, without a thou ght for yo ursel f." Because of actual or s uspected compromise, however, changes were made so frequently and compli cations so oflen introduced in the Ground systems that analysis became more and more difficult. By the e nd of the War the Ground problem reached the point where the time required for so lution made the production of currenl translations seemingly imposs ible. The general intelligence value and the special cryptanalytic inte rest of the problem, however, warranted studies of the last unso lved period of the highest-eche lon Japanese Ground Force system, for which no captured material or special cryptographic inte llige nce were available. These studies, success full y carried out during the final months of 1945, demonstrated that the development of c ryptanal ytic attacks had kept pace with the ever-increasing complex ity of Japanese cryptographic procedures. On Ihe other hand , no compromises or suspected compromises took place in the case of the Water Transport systems, and in the lalter, the Japanese moved in an orderly fa shion 10 make Iheir periodic changes. Consequently, the cryplanalysls were less hampe red by frequ enl or s udde n changes and a fair proporlion of Ihe messages in each key book were read cu rrentl y. Had no comprom ises been made in Ihe ground systems, the same success might ha ve been experienced. Therefore, it is a moot question whether compromised material gave an overall advantage or not. For lack of suffic ient personnel and because of the special interest of GCCS in the air systems, the SSA did nOI concern itself 10 any considerable extent wilh those syste ms until late in 1944. Thereafter, more and more attention was devoted 10 Ihe air problem and Ihe SSA eventually made large contribulions 10 currenl solution.

After some earl y compromi ses in 1944 the Japanese signal s systems were read from lime 10 lime. The cryptography used was such thai , with the limited volume of traffic, so lutio n would have been ex tre me ly diffi c ult wilhout a compromised code book, bUI fortunalely at different times several successive editions of the s ignals code books were caplured, a long with Ihe key books for a number of periods. The text of these systems, which discussed call -s ign freque ncies, and methods of handling Iraffic, was of special interest primarily to our traffic analYSIS. While Ihe stud y of low-eche lon Japanese syslems was never conside red a primary responsibility of the SSA, re ports from field agencies were examined here and Iheir conlenlS served as a guide in the training of mililary personnel deslined for fi e ld agencies. The SSA did aClive ly parli c ipate in Ihe solulion of Ihe low-eche-

108

SIGNALS INTELLIGENCE

Ion air syste m known as " BULB UL" but onl y as a suppo rt for the cryptanalytic unit in the India-Burma Theater. The impression has already been g iven that had more personnel been made ava ilable and at at an earlier date, solution of Japanese military communicat ions might have been expanded and exped ited. Yet it should be pointed out that had all the U.S. Arm y personnel wo rking on the Japa nese Army problems, not only at Arlington Hall Station but also in the Central Bureau at Bri sbane, in the Hawaiian Islands, and in the India- Burma Theater, been g rouped together at one center and solution activities thus been concentrated, considerable unnecessary dupli cation , especiall y in the field of translation , would have been eliminated. On the other hand it must be admilled that had such a consolidation taken place proble ms of ad mini strat ion wo uld have been great ly increased but the advantages gained by the increase of trained workers all applying their efforts in a coordinated allack wo uld have out weighed by far any administrative difficulties. As it was. where circ umstances permilled, duplication was eliminated and, considering the great distance between the agencies concerned, coope ration and coordination effected by the inte rchange of mail and telegraphic communications was good. In retrospect, it may be noted that in the signal intelligence field the consequences o f lack o f continuity and unpreparedness for effective operation immediately upon the outbreak of hostilities are nowhere more clearly demonstrated than in the case of the Japanese Army hi gh-eche lon secret communications. It is a fact that , during the entire period from 7 Decembe r 1941 to the summer of 1943, none of these communi cations was being read . Had thi s been otherwise, the military sit uation might have been quite different. To judge purely by the disastrous effect that the solutions obtained by us after the autumn o f 1943 had upon Japanese operations, it is legitimate to think that the important early Japanese penetrations to the south might have met with greater obstacles and that as a result the War in the Pac ific might have been te rnlinated many months ea rlier. If an adequate staff of c ryptana lysis had been e ngaged in stud yi ng Japanese Army traffic continuous ly from 1939, when the systems were solvable with comparati ve ease, complete continuit y could have been maintained from the ve ry outset of the War. After 1939 the systems became more difficult but neve r more difficult than they were in 1945 whe n. because of the possess ion o f a background of knowledge and expe rience built from successful reading of earlier pe riods, they were solved.

*** (4) German Army and Air Force Traffic. It has alread y been noted that a log ical di vision of work between the Briti sh and the U.S. Governments resulted in concentration of the SSA on solution of Japanese diplomati c communi cations and Japanese military traffic in the Pacific theater, leavi ng the exploitation of German and Italian military traffic in the European theater the primary responsibility of the

109

CRACKING THE CODES

British. [text withheld] In the tense days of early 1942, however, the advi sability of ins uring against the contingency that the British organization working on enemy communications mig ht be put out of operation by enemy action led to the establi shment of an SSA unit to serve as a back-up. Though the SSA played the minor rol e in the work on Gemlan AIll1Y and Air Force and Italian Amly and Air Force traffic, nevertheless it made noteworthy contributions. The Gelll1an Armed Forces employed two basic types of cipher machines. One of them, a modification of a commercial mac hine known as the Enigma, produced cryptogram s of a very high order of security, but faulty usage, Te utonic love of order, and addiction to ste reotyped modes of expression made it possible for the Britis h to solve a very large portion o f all the messages transmitted, yielding intelligence of the hig hest value. In thi s work the SSA served as a cooperating and assisting echelon, contributing new ideas, techniques, and machinery. As a result of excellent coordination of basic research and development with practical operation s, SSA engineers invented and built an electronic so luti on machine far in advance of anything hithe rto known for solving messages in the most complicated form of the Enigma machine as used by the Germans. In addition to the foregoing, s pecialy se lected messages were sent from England to Arlington Hall for study and solution by SSA special machinery. The necessity for speed brought into use spec ial communicatio n cha nnel s and there were cases wherein the answer to a s pecific problem was obtained by the SSA, wired back and in the hands of the GCCS cryptana lysts within 90 or some times as few as 60 minutes.

• •• The Gelll1an Alll1ed Forces also used a series of complex c iphe r machines for enc iphering teleprinter communications among the ir hig hest eche lo ns. He re again the SSA collaborated with [text withheld] in inventing and desig ning new machine ry as well as in testin g new techniques and procedures. Two machines in panicular, invented , desig ned , and built either e ntirely at Arlington Hall Station , or by an outside contractor working under the direction of SSA e ngi nee rs, Itext withheld]. .. were successfully employed [text withheld] In addition to making an important contribution 10 the victory in Europe, the ex pe rience the SSA ga ined in such collaboration will, of course, be very useful 10 the U.S. Amly in future research in this field. Mention must a lso be made of our contribution in the signal inte lligence operations in the M edite rranean Theater. Here again the SSA fumi shed key and basically-trained personnel for Signal Corps units working in co llaboration [text wi th held] on both Gennan and secret communications .

••*

110

SIGNALS INTELLIGENCE

(8) Commercial Code Traffi c. Exploitation of the information to be obtained from decoding traffic [text withheld I in public comme rc ial codes was carried on more or less continuously after earl y 1943. Thi s activity was at first confined largely to codes in the c hief languages of Europe but was ultimately exte nded to include a lso a group of Japanese commerc ial codes which provided a rich mine o f militarily useful information concernin g cond iti ons in the Far East. (9) Special Problems. In addition to the traffic already described, the SSA had to face a numbe r of other spec ial problems requiring other techniq ues. These inc luded the transcription of shorthand documents; the solution of open codes, a type of c ryptograph y in w hi ch a secret text is hidde n wi thin an oste nsibl y harmless message: the transcription and translation of "scrambled speech," that is, te lephonic and rad iote lephonic conversations in enciphered and unenc iphered form in foreign languages: and the sol ution o f secret ink messages. The last named type involved much work for the Office of Censorship as well as for MIS . In thi s fi e ld the SSA techni c ians accomplished feats not duplicated e lsewhe re : the recovery of printing on doc ume nt s which had been printed by use of inks soluble in water, In the case of two documents very valuable code mate rial s were recove red for the Navy; one involved a German, the other a Japanese code book. E. SOME GENERAL REMARKS The remarkable success which the cryptana lytic unit s obtained mu st not be a llowed to c reate the impression that a ny of the tasks was accomplished wi thout skill , training, patie nce, vigilance, and me ntal labor of the most ex hausting kind. A description of the essential features of a solved cry ptographic system may often seem simpl e and it may be imagi ned therefore that soluti on was easy, but thi s is rarely the case. Frequently, a simple cryptographic tri ck may be as difficu lt to detect as one of the more complex varieties; in cryptanalysis the effect of some minor complicati ng factor, inserted sole ly to prevent solution, may prove to be a serious srumbling block, though occas iona ll y it may a lso prove to serve in the e nd as the ente ring wedge leading to solution . The time and e ffort needed for solution , of course, vary with the system . A sim ple substitution ci phe r may require onl y a few minutes ' work by a sing le analyst; other systems may be so difficult that the e ntire efforts of a number of persons working for man y months are needed. Of the two basic types of c ryptograp hy. codes and ciphers. reconstructi on of the former is. as has a lready been noted , a slow, labori ous process, each code gro up having to be identified sing ly, and the large r the code. the longe r the time needed; solution of a diffic ult c ipher may take eq ually long but success is instantaneous rather than grad ual-at one mome nt the c ipher is unsolved, the nex t it is so lved. The reconstruction of a c iphe r mac hine is. of course a very long process, but whe n this is finished. the keys used in each day's traffic may st ill have to be so lved as they appear. For this reason. even whe n the

CRACKING THE CODES

III

machine has been reconstructed, reading or c urrent trafric may be delayed until enough traffic all in the same specific key has been received to permit solution, Systems differ so great ly that a counting of solved systems is no adequate indication either of the volume of work accomplished, or or the brilliance or the achievement of the cryptana lysts, During the War a group never numbering more than twenty, and for a long time many fewer, succeeded in making readabl e approximately twenty-rive systems, In the same period a group or about eighty pe rsons were needed for the solution of a single system, For thi s reason the e rficiency of the two groups cannot be eva luated in terms or the number or systems solved,

*** , , ,By August 1945, however, the trafric [tex t withheld] had been studied and tran slations were c urrently being prepared in large volume, The number or these translations is reall y the best ga uge or the cryptanal ytic achieveme nt s of the SSA, since they are the final products of the cryptanalysts, assisted by the combined e rrort s of intercept operators, clerk s, translators, editors, typi sts, and proofreaders""Factors causing Ouctuation of volume were comp lex: it is not always possible to ex plain declines, but on occasion the introduction of new cryptographic techniques had the errect or slowing up production temporarily until so lution was achieved, After cryptana lysis the messages in almost every instance still had to be translated-a sma ll percentage or the messages a re transmitted by the ir originators in English- and arter the translations were prepared in draft form , they had to be checked for diction , accuracy, and rormat , and then typed ror reproduction by the "Dino" process. " Master copies" of the tran slations were then once more proofread

ror accuracy of typing and the required number of copies pre pared, Following this, the y were forward ed at rrequent interval s during the 24 hours to MIS, In certain cases, where urgency required it, forwardin g was either by special courier or by electrical means, over spec ial cryptographic circuits, Even then, the task or the SSA was not rini shed, for it was necessary that the daily "SSA Bulletin" of translations be read carerully by personnel whose duty it was to compile there rrom voluminous inromlation which wo uld, in tum , be userul to the cryptanal ysts working in other problems, Thus the fruits or the labors or one work ing unit cou ld be at the disposal of any othe rs that might need it, In thi s connection it may be stated that the rree exchange or inrormati on and results among the traFfic analysis unit s, the cryptanalytic units, the translati on and " bulletin " units, the communications-security units, and so on, is absolute ly vita l to success in the whole cryptologic Field,

***

112

SIGNALS INTELLIGENCE

From A Brief History of the G-2 Section, GHQ, SWPA. In the aftermath of World War /I , the G-2 section of U.S. Armed Forces , Far East, commanded by General of the Army Douglas MacArthur, devoted some of its energies to preparing histories of its successes ill the war ill the Pacific. Included ill A Brief Hi story of the G-2 Section, GHQ , SWPA and Affiliated Ullits: Illtroduction to the Ime/ligence Series was a short history of MacA rthur's OWII World War /I COMINT organi:atioll, a co/laborator----and rival--{)f Arlington Hall . CENTRAL BUREAU (CB) Central Bureau, a combined Australian-American organization, was activated on 15 April 1942 under the command of Colonel Joe R. Sherr, one of the officers evacuated from Bataan. From its inception until the end of the war, the Bureau operated under the direction of Major General S. B. Akin , Chief Signal Officer of the Southwest Pacific Area, who was largely responsible for its establi shment and technical development. The first contingent of America n signals intelligence personnel to reach Australia after the outbreak of war consisted of six officers and eight enlisted men who comprised the 837th Signal Service Detachment. These men left the United States on 16 April 1942 on the last scheduled Clipper night to Hawaii. As all regular runs from Honolulu to Auckland, New Zealand, had been discontinued, they had to "sweat out" the remaining 7000 miles of their trip as space became available on USAAF and RAAF planes. From Sydney, they proceeded to Melbourne to join the GHQ establi shed by General MacArthur. Within the next two months the American contingent was augmented by 12 additiona l en listed men and 3 officers, including Major A. Sinkov who took over the command of the Detachment. The Australian Army component of Central Bureau was the Australian Special Wireless Group which had seen considerable service against the Gernlans in Africa and the Near East and which included some British personnel who had escaped from Singapore . The Au stralian Air Force component consisted of personnel assigned from Victoria Barracks, Melbourne. In the field at that time. Central Bureau maintained radio personnel as follows: AIF /A ustralian Imperial Forces/ - No. 51 WIT /Wireless Telegraphy/ Unit and the RAAF - No. I Wireless Unit. Attached to the latter as special instructors were three U.S. signal intelligence personnel (It. Howard W. Brown. Sgt. John J. Phelan . Sgt. Carl Card) who had been evacuated with Colonel Sherr from Bataan. The 837th Signal Service Detachment was deactivated in May 1943 and replaced by the Signal Intelligence Service (still a component of Central Bureau) under Col. Harold S. Doud, with Capt. Chester W. Ray as Commander of Troops. In the autumn of 1943. Colonel Sinkov became Commanding Officer, SIS, and

Durin g the 1920s, Herbe rt O . Yardl ey headed the legendary " Black Chamber," a secret cryptanalyti c bureau jointl y fund ed by the War and State Department s. Forced into retire me nt , he sou ght to rec oup by writing a sensational ex pose of Ame rican inte llige nce acti vities. (NSA)

Th is browns tone house in New York City served as the headquarters for Yardley's organi zation. (Loui s Kruh )

The brilliant cryptologist William F. Friedman became the first head of the Army 's Signal Intelligence Service (S IS). (George Marshall Foundation)

A rare photograph ti.lken of Friedman and hi s small SIS slaff around 1935. (lNSeOM)

From 1929 to 1942. the headquarters of the Signal Intelligence Serv ice was located in the Munition s Building. a large temporary structure built in World War I. (NA RA )

A member of the 2d Si gnal Service Company operates a radio direc tion finder o n Haw a ii shortl y before World War II. The compan y se rved as the co ll ec ti o n arm of S IS. (C.T . Du va l)

The battleship Ari:O/UI sink s under Japan ese attack at Peml Harbo r. The success or Ihe Signal Inlelligence Service in breaking the Japanese diplomatic machine cipher cou ld not rorestall a Japanese military surprise. (DA VA)

The perccived American intelligcnce failure al Pearl Harbor led top Army leadership 10 lind ways to bcller explo it the potentials of commun ications intelligence. C lockwise. Secre tary of War Henry L. Stimson. Assistant Secretary of War John J . McCloy. Assistant C hief of Staff for Int elligence Maj . Gen. George V. Strong. (NARA)

A graduation parade at Vint Hill Fanns School in Warren ton. Virgi nia . Successful pros· ecuti on of the signills intelli gence war demanded thousands of trained AmlY cryplOlogic specialists. (INSeOM)

A Japanese lang uage trainin g course condu cted at Arlington Hall SLation. Virginia. (lNSeOM)

A form e r women 's junior co ll ege, Arl in g ton Hall was take n over by the War Department in 1942 and becamc the new headquarters of the Signal Inte lligen ce Serv ice. (INSeOM)

Sta ff hea ds of thc Army Signal Sec urit y Age ncy pose for a group portrai t at Arli ngton Hall Station in 1944. ( INSeOM)

AmlY cryptanalysts at Arlington Hall Station work in an open bay in one of the large temporary buildings const ructed to meet the needs of the Signal Security Agency's growing work force. (lNSeOM)

Masses of ca lculatin g machines were used at Arlington Hall Japanese military traffi c. (lNSeOM)

(0

process and ex ploit

CRACKING THE CODES

11 3

Capt. I. W. Wolfe, Commander of Troops. Central Bureau expanded rapidly. Its stre ngth in 1943 was well over a thousand men and women. Technical direction unde r General Akin was vested in three assistant directors who we re the ranking officers of the three larges t conti ngent s: Co lo ne l Sinkov for the Americans: Lt. Col. A.W. Sandford for the Australian Military Forces, and Wing Comdr. H. Roy Booth for the Royal Australian Air Force. The Bureau brought under a unified contro l personnel from the signal intelligence serv ices of five nations. Included were the following military and naval forces: Army of the United States, Australian Imperial Forces, Australian Military Forces, Australian Wome n's Auxiliary Service, Briti sh Anny, Canadian Military Forces. Royal Air Force, Royal Australian Air Force, Roya l Australian Navy. Royal Navy, United States Women's Army Corps, Women's Auxiliary Australian Air Force, United States Army. The United States Navy was not included. In order to avoid dupli cation, promote smooth inter-service relationships, and assure the maxi mum benefit from the pooling of resources and centralizing of contro l, representatives of each of the 14 component services were detailed to work on all o f the major problems. Personnel were placed where they could work most effectively regard less of nationality or the component to which they be longed. Before the war came to an end the personnel strength of Central Bureau was more than 4000 and there were detachments in all parts of the Southwest Pacific Area. Of this number about fift y per cent were American personnel. Since the work of Central Bureau was always closely connected with GHQ, the organization followed General MacA rthur as he moved north ward. This took the headquarters of Central Bureau from Melbourne to Brisbane (September 1942), to Ho llandia (late summer 1944), to Leyte (Octo ber 1944). to San Mig uel (Luzon) (May 1945), and to Tokyo (Septe mber 1945). The first move came in September 1942, when GHQ was moved to Bri sbane. In Bri sbane the administrative control of the American contingent passed (February 1943) to USAFFE. Operational contro l. however. remained with GHQ throughout the war. In November 1945, its mission accompli shed and a fine record o f cooperati ve effort and notable achievement to it s cred it , Central Bureau was deac tivated. General MacA rthur commended the organi zation for its outstanding achievements and awarded the Legion of Merit to ten of its members.

From SRH 280. This il1lercepted message ./i·om the Japanese Amhassador to Na:i Germany. Baron Oshima, contains IIseful technical information ji"o/l1 inside the Third Reich, in elllding data on German experiments with jet planes and rocket bombs. The Goehbels referred to was Joseph Goebbels. the infamolls Na:i propaganda minister.

114

SIGNALS INTELLIGENCE

From: Be rlin (Oshima) To: Tokyo 26 July 1944 #756 A 4 part message complete. Urgent. In recent statements made by GOEBBELS as well as some that appeared in the Gennan newspapers and magazines, it is frankly admitted in their exp lanation of the European war situation that German weapons have to yield the palm to the enemy in both quality and quantity. However, at present the Gennans are striving to improve the quality and increase the quantity of their weapons, and there are not a few people who say that in time Germany will be able to make up for her present technical inferiority ' .... The Military and Naval Attaches have doubtless sent you detailed reports regarding this matter, but as [ thought it would be necessary in order to get an estimate of the war situation in general, I am sending you mere ly the following outline regarding the new German weapons. Moreover, although quite a number of announcements have been made already with regard to this subject , since there are some very important matters in thi s message relating to mili-

tary affairs, [ wish you would treat this message with particular care. I. Weapons connected with the Anny: (a) Tanks: I have already sent you information regarding the Tiger and Panther Tanks, but it seems that they are placing especial emphasis on the production of Panthers in large quantities . (b) Assault Guns: This gun is primarily a weapon to be used for the smashing of the enemy's pillbox machine gun bases in cooperation with the Infantry, but it is often used also in anti-tank warfare. Thi s is one of the weapons conceming which the Gennans are placing great emphasis at present so far as production in large quantity is concemed. It is an 88 millimeter gun (the same as the gun on the Tiger) and has a crew of four men. Part 2. (c) The Panther Terror:' This is a kind of rocket gun that can be used by two so ldiers, and is a weapon for close-range fighting to be used against tanks. (d) The Panzer Fist:' This is the same kind of gun as the above, but it can be used by one person. (e) Trench Mortars:" These weapons could hardly be called a new weapon today, but their size has been increased (up to 15 centimeters) and they are being used in large quantity. (I) Robot Tanks (Goliath):' This is a sma ll-type tank controlled by radio waves, and it seems that its effecti veness depends on the configuration of the ground. (g) TIle establishing of 311illery divisions wi th the self-propelled gun units acting as ~le main strength: TI,e object of such a division is to concentrate the use of 3I1illery. I wish that in paragraph (c) above you would amend the words Panther Terror

CRACK ING THE CODES

11 5

to read Panzer Terror. 2. Weapons connected with the Navy: (a) A New Type Submarine (Very Secret): This new type submarine scarcely sho ws itself on the surface of the sea. and has long-range maneuverability. At present it is in process of construction and its effect wi ll first be felt from abo ut September, and by November or December it is said that greater results will be effected by its use in great quantity. Part 3. (b) Torpedo Boats: I have nothing to add to what I have already reported to you concerning these boats. (c) A Small-type Submarine (Very Secret): This is a submarine that has been improved one degree and seems to be a weapon that has two torpedoes to be ridden by one man. It is to be used in areas ncar the seashore. (d) Special Torpedo (Very Secret): This is a torpedo wh ich by means of special equipment has been made almost 100% effective in hitting the mark, and its testing has already been completed. (e) The High Speed Boat:' This in it self is really not a new weapon, but a new type is being produced wit h a speed of 40 knots an hour, and having a Diesel engine. For this reason it has the spec ial merit of not easi ly catch ing fire when damaged. It seems that the Gernlans are putting forth efforts to produce this boat in large quantity. (0 Special High Speed Boat (Very Secret): This is a high speed boat to which has been added the principle of radio wave control , and it seems that it s explos ive power is especia ll y great. 3. Weapons connected with the Air Force: (a) The Gennan a re adding substa ntial improvements to the speed, annament, amm unition, and way the armor is put on, etc., in their va rio us types of fighter and destroyer' planes. and besides that they are of course placing spec ial importance upon the prOduction of these planes in great quantity. Furthennore, in the near future propellerless planes are also going to make the ir appearance, and it seems that these planes wi ll have especia lly high speed (this last paragraph is very secret). (b) For the time being dive bombing has been suspended and as a subst itute they arc using ground attack planes (SCHLAC HT FLiEGER). These planes cooperate with ground fighting, and considerab le results have a lready been achieved in the destruction of enemy tanks (Captain RUUDERU alone, it has been announced, has destroyed more than 200 tanks). Pan 4. (c) The main strcnth of anti-aircraft gu ns lies in the 88 and 105 millimeter guns, it seems. However, 120 and 150 millimeter guns also are gradua lly being used in large quantity. MatChing the improvement in shells the Gemlans are trying also to increase the power in their ai rcraft Again it seems they are also using rocket guns.

SIGNALS INTELLIGENCE

116

(d) There has been a great deal of improvement in night fighters since last winter but in day-light air raids also special units, STORM GRUPPE, have recently been organized (within 2 minutes I American four-motored plane- IOG- and -M-). They wi ll at an appropriate time be newly established and although it may not be possible to call them new weapons, in outline they serve as an indication of the earnest efforts which Germany is puting fonh in order to regain command of the air. (e) Special Type Aircraft (Very Secret): It seems that a study is now being made with regard also to a bomber plane which will unite the rocket gun with the principle of radio wave operation. 4. Equipment for Long Distance Bombardment (Very Secret): I have already described to you in various telegrams the V-I. and it seems that preparations have already been made for the use of a more effective weapon than this. It will have five times the effect of the weapons thus used in the past. And again it is said that they will have a weapon whose effect will extend to a square of more than I kilometer. However, it is not clear as yet just when both these weapons will be introduced. It is anticipated that with regard to these weapons there will be both substantia l developments and improvements made in their construction.

5. The weapons in which the Anglo-Americans made particularly great progress in the past were first of all the direction finder and then radio wave weapons in general, but in both these spheres the Gemlans by specia l efforts seem gradually to be catching up with the enemy. It seems that in weapons connected with rockets, because of the fact that the Gennans as usual have had longer experience, Germany is temporarily in the lead. Again improvements and progress in explosives have been best indicated in the V-I and it seems that this improvement has been very conspicuous. a-Y.H.1891.

b- SCHRECK. c - The Kana spelling for this word is HUAASUTO, which is doubtless the Gemlan word for FAUST. d - WERFER. e - The Kana spelling for this word is GORIAATO. r - SCHNELL BOTE. g - ZERSTOERER.

Inter 28 Jul 44 (I) Rec'd 29 Jul44 Trans 3 1 Jul 44 (1227-A)

Japanese

CHAPTER VI Producing Intelligence The decryption of encoded and enciphered enemy traffic was only the beginnings of the s ignals intelli gence process . The messages had to be placed into a wider intelligence context. Until June 1944, when there was a general reorgani zation of the whole War Department intelligence structure, this was done by the members of the Special Branch, Military Intelligence Service. In the excerpts from SRH 185, War Experience s of Alfred McConnack, the father of the Special Branch gives an insider's account of the infighting that went on within War Department inte lligence. Unlike mo st of the other Specia l Resea rch Histories relea sed by the National Security Agency, thi s document was originally prepared after the war as an unclassifed study; the reader will note that McConnack does not mention that the Special Branch dealt with signal s intelligence. The Special Branch not only exploited U.S . Army successes against Japanese diplomatic and later military communications, but also entered into an agreement to share signa ls intelligence with the British. As a result, after 1943, the Uni ted States was able to sec ure from the British the results of their unparalleled success in reading hi gh- level German messages. See the excerpt from SRH 141-2. The type of intelligence windfall that resulted from this so urce is documented in the excerpt from SRH 005, Use of MX/CSS ULTRA by the War Department.

118

SIGNALS INTELLIGENCE

From SRH /85. /n Ihese excerplS from his unpublis hed m e moirs, McCormack lells all aboul Ihe Special Branch---excepl wh(1I il aCll/ally did! "Crava/h melhods" refer 10 Ihe highly seleClive and personalized recruilmen I procedures used by his old lawlirm. AI Ihe lime he wrole Ihis aCCOUl1l, McCormack had rell/rIled 10 a successjitl law praclice. Afler reel1lering civilian life in 1945, he had briejly served as Ihe Slale Deparlmel1l 'S firsl DireClor of Il1lellig el1ce and Research. Howe ver, he apparel1lly ellcolllilered Ihe same kinds of bureaucralic resis/allce 10 inlelligence work al Ih e Slate Deparlmel1l as he lells us he mel in Ihe War Deparlmenl . 3 1 July 1947 WAR EXPERIENCE OF ALFRED McCORMACK Early in January 1942, McConnack was asked by McCloy to do a job in the War De pa rtment in connection with military inte ll igence, and he took office on January 19 as a Special Assistant to the Secretary of War. In March he submitted hi s recommendations to the Secretary and was given the job of putting the m into effect. A special section was created by G-2 to assist in the work, and in May 1942, Colonel (later Brig. General) Carter W. Clarke, with whom McConnack had become acquainted and whom he strongly recommended, was made head of that section. Urged by the Assistant Chief of Staff G-2 , General Strong, to go into the Army, McConnack at first declined but late r concluded that hi s job could be done more e ffec tively if he were an officer on the G-2 staff. He was comm iss ion ed as a Colonel on June 24, 1942, reve rsing roles with Clarke, who the re upon bccame Chief of a newl y created "Special Branch" of G-2 , with McCormack as Deputy Chief. From the n until the e nd of the war McCormack and Clarke, whose talents were compleme ntary, worked as a team. McCormack took charge of the research and intelligence work and the recruiting and training of personnel, while Clarke concerned himself with admini stration , security, relations with G-2 and outside organizations and innumerable proble ms involved in organizing and building up units of the Anny whose work produced the raw materials of intell igence, and for which Clarke ass um ed a continually increasing respon sibilit y as the war progressed. Clarke was a lso e mplo yed on a numbe r of important mi ssions for the War Department Staff. Both Clarke and McConnack conside red G-2, as it ex isted in 1942, 10 be inadequate for Ihe inte lli ge nce job that the war required , and they soug ht to use the Special Branch as a means of building up a more effecti ve intelligence organization in the War Department. The original assignment of Special Branch was the preparation of a daily intelli gence re port , but the work which that e ntailed provided man y opportunities for

PRODUCING INTELLIGENCE

11 9

the prod uc ti on o f useful intelligence, whi ch the pe rsonne l of the Branch were quick to grasp. In August, 1942, McCormack. hav ing establi shed the foml o f the report and the methods to be used in the preparation, turned over the work to He nry W. Ri gby, who had been a Cravath associate from 1937 to 1939, before going to the Prov ide nce firnl of Tillinghast, Collins, & Tanner, and who had fo llowed McComlack into the War Department in March 1942. Ri gby's immediate assistant until August, 1944, was a C ravath associate, Louis T. Stone. Jr.... McCormac k used "Cravath me thod s" in the recruiting of personnel for the Special Branch , taking on only hig hly qualified pe rsons. There were so man y lawyers in the Special Branch that it used to be referred to humorous ly as " the best law office in WaShington". McCormac k be lieved, and hi s ex perience confirmed, that lawye rs as a class are bette r fitted ror inte lli gence work of the type that must be done in the War De partment, i.e .. what the Anny ca ll s "strategic intelligence", than is any other gro up in the community, because their training teaches the m to deal with evidence, to be inquisitive and skepti ca l, to pursue an investigation through to a concl usion, to meet unfamili ar situations of fact and, most impo rtant , to do very detailed work without losing their sense of va lues. After rreeing himself from responsibility ror daily reporting, McCormack turned hi s atte ntion to spec ific problems of inte llige nce, being convinced tha t intensive and ex haustive research work would furnish de pendable answers to man y questions on which G-2 at the time cou ld do no be ller than speculate. Every job that was done revealed new opportunities and emphasized the need for more personnel , and the results were good e nough so that the personnel allotments were provided, though usually after long delays and much paper work. McCloy had an active interest in inte lligence, followed the wo rk o f the Special Branch and he lped out freque ntl y; and the fight for personne l was made easier by high prai se which the Branc h received from time to time from General Ma rsha ll and other high officials, and occas ionally from the President. By the Fall of 1942 it was apparent to McComlack. and a lso to Clark, that G-2 had no dependable curre nt inrormation about the war in Europe. The reason was a s impl e o ne. but nobody seemed to rea li ze it. There was a m ilit ary attache in London wit h a large starr. He sent reams of paper back to WaShington, whic h he had gathered from the British intelligence agencies; bu t the British did not regard military all ac hes as being entitled to top leve l information and so what the attache office sent home was seldom mo re than could be published in the newspapers. It was so lemnl y studied in G-2, though there was in the organi zation an a lmost unive rsal suspic ion or the British and a tendency to disagree with British conclusions, si mpl y because they were Briti sh conc lu sions. Some of the G-2 " inte llige nce" on Euro pe was rantastic. For instance, at a time whe n the German Air Force had about 5,000 operational planes or all categories. there were devout believers in G-2 in a rumor picked up by an attache in Roumania that the Germans had 10.000 spare airplanes hidden in the Black Forest.

120

SIGNALS INTELLIGENCE

One of the bothersome characteristics of G-2 was a certain supine attitude toward intelligence. Everybody knew the classical Army doctrine that the three steps in intelligence are collecting, evaluating and disseminating, but nobody seemed to give much thought to the fact that evaluation and dissem ination a re worthless if what is collected is worthless. The attitude was in part a reflection of the viewpoint of the G-2 colonels, who regarded it as their function to make predictions. They operated in the manner of soothsayers. They showed no indication that their infortnation might be inadequate. They seemed to think that they already knew enough to answer whatever questions might be put to them concerning the enemy countries.

G-2 organization as a whole did not appear to know that there was a war on. Except for the fact that they were required by departmental order to work on Saturdays and that some of them had to appear Sunday morning for a meeting with the General, they operated their organizations just as they had in peace time. Promptly on the dot of 5 p.m. the who le organization went home. There was no special sense of hurry or urgency and the job that did not get done one day would go over to the next unless the job had a deadline. If the job had a deadline, it was turned out on that deadline and it was good or bad, depending on whether the time was su fficient to get the job done. The Special Branch had a different attitude toward hours. Every day's work had to be done that day and if it was not done by 5 p.m. the personnel stayed on and finished il. The Branch was hardly ever closed before II p.m., and it opened very early in the morning because Clarke's hour of arrival was never later than 7 a.m. and he not infrequently got in at 6. McCormack never went home before 7 and worked at night whenever there was any reason to do so. McCortnack had an early experience which he never forgol. At 4:40 one afternoon he went to the Chief of the European Branch of G-2 to get some information. The campaign in North Africa was then under way and the infonnation, though not of any vi tal significance, was pertinent to a job thallllight have been useful in connection with that campaign. The Colonel apologized for not being able to get the infornlation until the following morning. It was a fixed rule of hi s branch, he said , that all personnel must s tart putting their papers away by 4:40 p.m., in order that their desks might be clear and might be inspected and the files locked by 5. In the two years of life of the Special Branch it g rew in size from a dozen officers and civilians to a total of 100 officers and 300 civilians and enlisted personnel in Washington, and 80 officers abroad, stationed with the field commands and at other key places, and it had been g iven a directing responsibility for the work of ArnlY units outside of Washington numbering several thou sands of persons. McCormack's work in early 1943 had to do largely with the war in Europe. At that time the relations between the War Department and British intelligence organizations were conduc ted at the top level; there was no working li aison and no important !low of the raw materials of intelligence from the British organizations

12 1

PRODUCING INTELLIGENCE

to the War Departme nl. In April , McCormack was sent to Eng land as the represe ntative of the Assistalll Chief of Staff G-2, to work out arrangeme nts thai wou ld insure that the American fie ld commands would be furni shed promptly with all relevant Briti sh intelli gence, and that the War De partme nt would have access 10 British intelligence in all its aspects. McCormack spent two months making a stud y of the British Intelligence agencies and reporting hi s findings to Washington , with recommendations for methods of coordinating British and American intelligence work on the European wa r. He was greatly impressed by the work of the British intelligence agencies and the high quality of their personnel, and he learned many lessons from the m which he applied later in G-2. He found that Briti sh intelligence had a first call on military and civilian personnel-a startling contrast to the United States, where intelligence had in general a low priority for personnel or no priority at all. McCormack and LI. Col. (later Colonel) Telford Taylor, ... who had accompanied him to England, worked out with the Briti sh a plan for coordinating Briti sh and American intelligence work for the German war and it was gradually put into effect as personnel were obtained. It invo lved establi shing a unit of the Specia l Branch in England, headed by Colonel Taylor, converting certai n British intelli gence units into joint British-American units, and creating a new system of communications for the transmission of operational intelligence to American commanders.

*** On his return from England in June, 1943 McCormack set up a new unit under Cap I. (later LI. Col.) William R. Perdue of the Cahill, Gordon finn and LI. (later Major) John W. O'Boyle, a Milwaukee lawyer, who had se rved as Jud ge Patte rson's law cle rk in the Circuit Court of Appeals and, before go ing into the Army, had he ld a number of important assignments in th e Le nd-Le ase Administration. The purpose o f the unit was to expl o it Japa nese intelligence mate rial with primary altention to the Japanese merchant marine. No two intelligence organi zations among the Allies could agree at that time on the amount of Japanese shipping at the start of the war or the amoulll of tonnage that had been sunk , and there were estimates of Japan's shipping position that varied by millions of IOns. Clarke and McCormack were agreed that the Japanese merchant fleet was the primary military object ive in the Pac ific War and that the emp loyment of that fleet was the best clue to the logistic problems of Japan and to her economic situation and military capabi liti es. Seeing an opportunity to do an importalll intelligence job in that field they requested, and after some months of de lay obtained, SS officers and 22 en li sted WAC's for the Perdue-O ' Boyle unil. The job was compl icated by language problems. Not only was much of the intell igence material in Japanese, but the pre-war ship li sts were a riddle because

122

SIGNALS INTELLIGENCE

the Japanese had an annoying habit of g iving the same name to a number o f ships (42 ships in one instance) and an eq ually annoying habit of rendering a single ship name into the kana sy llabary in several different ways. Experts in the Japanese lang uage we re needed , and th ere was a great shortage of the m . Again app ly in g "Cravath methods", McConnack persuaded Harvard to re lease Professor Edwin O. Reischaue r, its Director of Japanese Studies, and with Re ischauer's assi sta nce recruited the best staff of Japanese linguists in Washington. The intelligence work on the Japanese merchant marine, which later became a joint Anny-Navy operation unde r McConnack 's supe rvi sion, was slow in getting unde r way, because of the perpetual difficulty of getting personnel;, but it proved to be excepti onall y successful and contributed largely to the effectiveness of the submarine war against Japan. The United States Strategic Bombing Survey fo und , when they went to Japan afte r the wa r, that the infol111ation on Japanese shipping that Washington had had during the last year of the war was far more detailed a nd accurate than anyt hing that could have been put together in Tokyo at the same time. From 1943 to 1945, McCormac k was a member of the Joint Army-Navy Assessment Committee, which had the res ponsibility of detennining e nemy shipping losses. The intelligence work on the Japanese merchant marine led into many other important subjects of Japanese intelligence. Techniques established in that wo rk became applicable to other intelli gence problems affecting Japan; and the application of those teChniques accounted in large measure for the graduall y increasi ng accuracy of American intelligence on the Japanese military and economic situation. The work led, almost necessa ril y, into wo rk in the Japanese Order of Battle unit and a gro up was set up to do that work para lle ling the work of the regular Japanese Order of Battle unit of G -2 and dividing the responsibility with it. The Special Branch reported directly to the A. C. of S. G-2, while the othe r intelligence units of G-2 were one echelon lowe r, under the Chief of the Military Intelli gence Service. With some exceptions, they were organi zed on a geog raphic basis, and the geographi ca l sub-sectio ns had become highl y compartme ntali zed. The Special Branch and the Order of Battle were strictl y war products and confined their att enti on to inte lli gence tha t was relevant to the wa r, a nd mainl y to inte lligence of combat significance. The Geographical Branches , o n the o ther hand. had taken their f0l111 and acquired their habits in peacetime. They attempted to cover the who le world of intelligence; and their contact with combat intelligence was often rather remote. In the upper echelons of G-2 there were some ab le officers, but most of the regular ones knew that their futures depended on their gett ing assignme nts in the field, and of those who were wi lling to remain , because they realized the importance of good intelligence work, many became discouraged and obta ined overseas jobs. Their re placements were often men wi th no better qualifications than that their shoulder insignia fitted the job; and thus from 1942 to 1944 a law of natural

PRODUCING fNTELLlGENCE

123

selection worked against G-2 adversely. The discouragements that made able officers get out of G-2 were numerous. There was a lack of imaginative leadership on the "Colonel" level. There were jealousies and political fights within the organization. There was an undue emphasis on rank and an attitude of unwillingness to give responsibility and credit to subordinates. There was a certain feeling of remoteness from the war. But the greatest discouragement of all was the inability of able officers to get qualified personnel. and in a sufficient number, to do their jobs. Difficulties in getting personnel were partly due to a lack of recognition of the personnel needs of intelligence, and even a lack of appreciation of the importance of intelligence, on the part of the War Department General Staff; partly to the politically motivated drive to keep the number of officers in Washington to a minimum; partly to the Army's heavy-hand methods of handling personnel and partly to the stupidity of the Civil Service organization. Though intelligence work requires a higher order of ability, War Department intelligence had practically no personnel priorities until McCormack got some priorities established in 1944. Moreover, intelligence personnel were subject to sweeping orders from above, which came unannounced from above and invariably caused trouble. [n one of his early efforts to get civilian personnel, McCormack and his staff had interviewed dozens of people and had selected some 15 or 20 highly qualified ones. These were put through the long process of security clearance and Civil Service approval. Then, just as the clearances and approvals began to come through, a "personnel freeze" for the War Department General Staff was announced. No more persons could be hired until further notice. By the time the freeze was lifted practically all of the candidates (their security having been cleared and their Civil Service status established) had taken jobs with other agencies, and the recruiting of personnel for the vacant jobs had to be started anew. The civilian agencies of government were hiring personnel at somewhat inflated salary scales, and under war-time conditions highly qualified persons could not be drawn from useful jobs in private life unless they were paid enough of a salary to support their families. But the G-2 salaries were very low. They were paid with one eye on the base pay scale of officers, and officers found it difficult to understand why a competent economist had to be paid $8,000, the base pay of a Brigadier-General. When McConnack went into G-2, the highest paid civilian in the Europe-Africa Branch of G-2 got $3,600 per year as salary . He was the "expert" in German economics; and it was only to be expected that he would beas he was-unqualified for the job. The recruiting of officers was very difficult. They had to be of ranks that fitted into a Manning Table. They could not be taken from either combat units or specialized schools. The ArnlY had no OCS for intelligence personnel. Much of the work in McCormack's unit was young men's work. requiring much enthusiasm. long hours, and great physical and mental energy. So he tried to find bright young offi-

124

SIGNALS INTELLIGENCE

cers, and did find some ve ry good ones. Then came an order forbiddin g the assignment in Washington of any more officers under 28 years of age, and somewhat later an order requiring all officers under 28 to be sent to the field. Enlisted personnel were not pemlitted in G-2, except for a number of men who served as guards at night and in similar jobs. To get an enli sted man it was necessary to have him commissioned; and McCormack was able to get that done in a few cases, but it was a difficult and time-consuming job. The most maddening difficulties came from the Civil Service, which cou ld not understand why the war should interfere with it s time-worn procedures. If G-2 wanted to hire an indexer, it was necessary to prove to C ivil Service, not only that the candidate was qua lified, but that an indexer was really needed. Thus Civil Service exami ners, many of whom were incapable of understanding what G-2 was do ing or what its problems were, quite o ft en turned down an application on the ground that the work was not necessary.

*** Major General George B. Strong, the Assistant Chief of Staff G-2 , attempted to give G-2 a vigorous and prog ress ive administration. He encouraged and helped the Special Branch, and he recognized Colonel Lovell 's work by creat ing an Order of Battle Branch with Lovell as Chief and centering in it all work on enem y Order of Battle. He encouraged another Branch of G-2 which proved to be of great value in the war-the Topographic Branch, headed by Colonel S. P. Poole, a geographer from the University of Rochester. But General Strong was in failing health and was habituated to a system in which an officer must accomplish his results with the personnel given to him- an attitude that prevented him from seeing that what was needed in G-2 was a thorough shaking up. As one junior officer put it, the quickest way to refoml G-2 was to kill off all officers above the rank of Major, and to throw in half of the Majo rs for good measure. The career of the Special Branch was largely shaped by the deficiencies of G2, in that it had to take on job after job that G-2 should have been doing but was not doing . McCornlack discovered , even before he went into the Army, that G-2 had no biographical files for identifi cation of foreign indi vid ual s. He proposed to the G-2 Chief of Intelligence that the latler set up a central name file, in which there would be assembled all information about persons figuring in G-2 intelligence material and for which a group of research wo rkers wo uld gather additional material from biographical sources. After some months, hav ing got no action from

G-2, McComlack decided to set up such files himself. He assigned a ve ry competent woman ... to do the job and , being unable at first to give her any assistants, gave her instead a call on the free time of stenographers. The unit grew steadil y, as it proved the importance of foreign biogra phical infor-

125

PRO DUCING INTELLIGE NCE

mation. By 1944 it had become the " Who's Who" branch, with a staff of over 180 persons, headed by Lt. Col. Edward C. Lapping, now Managing Editor of the Chicago Herald-American, and it had become a primary source of informati on abou t foreign indi viduals, serving not onl y the War Department but all other govemmental agencies requiring such info rmation. McCormack tried to keep the Spec ial Branch a llexible organi zation, so that, if a job of press ing impo rt ance came alo ng, as many of the starf as were needed could be thro wn into it. The G-2 organi zati on, on the other hand , was rigidly organized along lines or command, do wn to the highl y compartmentali zed geographi cal "desks", and there was no flow of personnel from one "desk" to another to meet emergencies. Late in 1943 Colonel Moses W. Pettigrew, Chier of the Far Eastem Branch, whom McCormack considers to be the ablest American inte lligence officer on Japanese matters, was wo rki ng up a Japanese Order of Battle, wi th the assistance of an able younger offi cer, Lt. Col. (later Colonel) Eric H. H. Svensson and a woeful y inadequate sta rf. Co lone l Pettigrew had to do most of hi s wo rk at night, as du ring the day his time was consumed in the admini strati on of a cumbersome organization, in routine assignments from his superiors, and in a largely fruitless effort to get competent personnel. In spite of the handicaps, Colone l Pettigrew laid the foundations for work on the Japanese Order of Battle and was the Arm y's leading expert on the Japanese AmlY.... An experience of Colone l Pettigrew's illu strates the di ffi culties that the able G-2 officers had to contend with. Whil e he was building up hi s Order of Battle, a priceless inte lligence doc ument was captured and was translated in Australia- a Japanese AmlY Office r Li st of comparatively recent date. Here, if the li st could be carded and the cards kept up to date with all information coming in abo ut Japanese Arm y officers, was an extremel y useful too l for Order of Battle inte lligence. Pett igrew applied for personnel to do the job--two Japanese ling ui sts and 6 ty pists. The application was turned do wn. Out of the hund reds of persons in G-2, no one could be spared fo r Petti grew's job. Petti grew appealed to McCormack, who saw the impo rtance of the job and said he would get it done somehow....

*** Earl y in 1944 the work of the Special Branch had grown and developed to a point where a much larger staff was required to take ad vantage of the inte lligence opportunities that were c learl y there. The personnel authorities, however, had set an overall ceiling on G-2 personnel in Washington, and it was ruled that no future addit ions of personnel ror the Spec ial Branch could be made except by transfer of jobs fro m the rest of G-2. There were so many vested interests in G-2 that the war would be over before any considerable nu mber of persons could be obtained that

126

SIGNALS INTELLIGENCE

way; and so it looked as if many promising opportunities in intelligence opportunities would have to be passed up. It was Clarke who at that point came to the rescue with a bright idea, as he often did. After all, it was Secretary Stimson who had brought McCormack to Washington and given him a job to do in intelligence. Two years had passed and the Secretary had not yet asked McCormack how he was coming along. It was high time for the Secretary to ask that question, and no doubt he would ask it if he learned that McCormack was in trouble. Clarke was right. The Secretary requested General Strong (who had requested him so to request) to have McCormack prepare a full report. McCormack did so, accompanying it with an urgent application for an increase of officers from 100 to 400 and of civilian personnel from 300 to some 800, plus cel1ain additions to personnel abroad. The result was appointment of a War Department Board to pass on the application. It consisted of McCloy, the Deputy Chief of Staff (General McNarney) and a number of other general officers. The Board heard evidence and argument and found that the personnel applied for could be profitably used. It granted the requested increase in the TIO, but it limited the addition of new officers to 100 pending an investigation of the entire G-2 organization, to determine what compensating personnel reductions could be made in other units. An investigation of G-2 by an Inspector General's Board was then ordered, and Clarke and McCormack spent several weeks explaining their organization and plans to the Board members. The Inspector General's investigation resulted in recommendations which summed up to an abolition of the geographical organization of G-2 and the merger of those units and the Order of Battle Branch into the Special Branch. A new Board was then convened for the purpose of devising a suitable plan for the reorganization of G-2. The plan was adopted on or about April I, 1944 and a new Assistant Chief of Staff, G-2, Major General Clayton Bissell, was directed to put it into effect. General Bissell thereupon appointed a Reorganization Committee to translate the plan into specific terms and promptly departed for a trip abroad, directing the Committee to have its recommendations ready on his return. The members of the Committee were either old-line G-2 officers or officers whom General Bi sse ll brought in from the outside, most of them having no experience in intelligence. The plan of reorganization divided G-2 into a Policy Staff to advise the G-2 on weighty matters and the Military Intelligence Service to do the intelligence work. The Military Intelligence Service was in turn divided into a Director of Administration (to run the service organization), a Director of Information (to take charge of procurement of information and to run a variety of unit s engaged in that activity) and a Director of Intelligence to do the intelligence work and to provide intelligence for the War Department and the commands abroad. Under the Director of Intelligence there were to be a group of Area Specialists

127

PRODUCING INTELLIGENCE

without staffs, a Research Unit and a Reports Unit. The bulk of the organization was to be in the Resea rch Unit, which was to be divided into the following Branches: Military, Political, Economic, Sociological, Topographic , Scientific, Who's Who and Library. The Reorganization Committee was dominated by the old-line prejudice toward the Special Branch and made it a primary objective of their plan to ensure that the Special Branch would be broken up-"and good". The plan distributed its personnel and functions among 53 separate branches, sections and sub-sections, and the literal provisions of the plan would also have destroyed it s work. The key position in the new organization was that of Director of Intelligence, and a strenuous attempt was made to find a suitable regular officer for the post. The attempt was not successful and, largely through the efforts of Clarke, McComlack was appointed to the post. Clarke was appointed Deputy Chief of the Military Intelligence Service. Brigadier General Russ Osman, Regu lar Army officer, was appointed as Chief. General Osmun, who belonged to the Quartermaster Corps, was replaced some months later by Brig. General Paul E. Peabody, who had returned from service as Military Attachc' in London. General Peabody gave Clarke and McCormack the fullest support and did much to help them get their jobs done. The reorganization was gradually put into effect in June, 1944, and it attained some newspaper fame by reason of the fact that the Sections doing Gennan intelligence were moving their offices, and were without telephone connections. on the day of the Normandy inva s ion. That was of a piece with the rest of the Reorganization Committee's work. It set up about as impracticable a scheme of organization as could be devised. It apportioned responsibilities among the 3 units under the Director of Intelligence in a manner calculated to render all three useless. It put into key positions a number of officers without a shadow of qualifications. All McCormack's subordinates, down to the Branch Chiefs. were chosen for him and without consultation with him. and the personncl were di stributed among the Branches according to the fancies of the Committee. McConnack tackled his new job with a determination to weed out all misfits and put every job in the hands of a person, officer or civilian. clearly qualified to do it. He started out on the day when he took office by demanding the removal of the Acting Supervisor of Rcports , on the ground that he was incompetent and dangerous. With the support of Clarke, and by submitting his resignation in writing, McCormack got the ofticer removed before he had takcn office. Then followed 4 months of agonizing efforts to make an organization ou! of the formless thing which the Committee had set up.

***

SIGNALS INTELLIGENCE

128

From SRH 141-2. This excelpt from another McCormack memorandumagain . dealing with personn el--dis c nsses the backgronnd of the 1943 British-War Department agreement that at last let the U.S. Army in on the ULTRA secret. 23 Oc to be r 1943. MEMORANDUM FOR MR. McCLOY (Through GENERAL STRONG ): Subject: Personne l Situation of Special Branc h, M.I.D.

*** 30. German .. £". The most important deve lopme nt affecting the Special Branch, since April , has been the opening up of the German "E" traffic. Thi s has been the No. I source of military intelligence during the present war, from ea rly 1940 to date. The achieveme nt of the British in thi s fi eld has been without question the greatest feat in the history of cryptanalysis, and the preservation and increasingly successful exp lo itation of this source for three and a half years has been an amazing achievement in security. 3 1. From the latter part of 1942 until May 1943, a controve rsy existed between the War De partme nt and the Briti sh GC & CS, as a result of the refusa l of the British to make Ihis traffic available 10 the War Department in any way, and their unwillingness to consent to Arlington Hall 's entering the "E" field. An ag ree ment between the War Department and the GC & CS was worked out by Gene ral Strong in May, pursuant to which thi s source of intelli gence was to be opened up. The agreement ca ll ed for (a) es tabli shm e nt of a n American cry ptanal yti c un it in England. to work o n "E" in coope ra tion w ith the Briti sh. (b) assignment to Arlington Hall of a research job in the "E" field and (c) opening up of the traffic it se lf to the War department through liai son officers in Eng land and the Special Branch in Washington. 32. Because of the g reat ri sks invo lved in the dissemination of intelli gence from "E" sources. and the necessi ty o f sati sfy ing the British that ou r sec urit y arran gements arc adequate-and the necessi ty also of persuading them to agree to

our wishes as to what "E" intelli gence should be tran smitted to Washington- it took from May to the end o f August to work out details of the arrangements with the GC & CS in the ''E'' field. These call for 4 offi ce rs to sit on shift s in the socalled " Watch Room" at B.P., where the material comes in from the cryptanal ysts and is eme nded. translated and disseminated (a) in origi nal form to the Service Ministries by teletype (b) in the form of informational messages to comma nders abroad and (c) in memoranda for the Ministries which gat he r up and present the

PRODUCING INTELLIG ENCE

129

info rmation thai is not of sufficient imporlance o r urgency to go through c hanne ls (a) and (b). 33. In addition, one liaison officer is to sit in the War Office, at the point where "E" intell igence comes fTOm B.P., and is put together with other intelligence into "appreciations" for the War Office and commanders abroad. This officer (who is being taken from another liaison assignment with the War Office) is to advise our liaison offi cers at B.P. what additional or explanatory material should be sent to WaShington. 34 . Alle nti o n is ca ll ed to th e fac t th at, unde r thi s arrangeme nt, the Wa r Deparlment gets, for the first time, a basis for arriving at an indepe ndent judgement on British inte lligence from "E" sources, which is the majo r part of British intelligence about the Gennan Arnl y and Air Fo rce. He reto fore we have depended largely on the infe rences and conclusions d rawn by the British, and have had no means o f checking their basis. We have been in the same position as those organi zations and indi viduals within the British Government who are not Hin the know," i.e., the MS in fonnation has been eithe r wi thheld or transmilled to us late, for securit y reasons; or it has been given in general terms and in round numbe rs; and ofte n it has been impossible to di stinguish betwee n 100 % ve rified info rmation and infe re nces which, upon examination of the ev ide nce, might tum out to be mere g uesses. 35. Unde r thi s arrangeme nt we need no longer spec ulate a bo ut how muc h intell igence the Briti sh are g iving us from their primary source. We now sit at the source itself and can watch it ; or rathe r. we will be able to watch it whe n we get the necessary component of officers at B .P. 36. The Briti sh Security channel to Was hington is used fo r send ing the HE" messages and paraphrases here, and gives us d irect and speedy access to the intelli gence deri ved from HE" and related c ryptanalytic acti vities. The nature and importa nce o f this intell igence can be dealt with orally, and specimens of it can be furni shed. Because of the unique securit y problem invo lved, it is undes irable to deal with th em in a memorandum.

37. T he above arrangement also gives Ame rican officers a pari in the dete nnination of what HE" intelli gence shall be sent to Ame rican fi eld commande rs. T he disseminati on of HE" inte lli gence to the fi eld is controlled at B.P., whe re re presentatives o f the 3 Serv ice Min istries dec ide what their respecti ve commanders should get, and in the past they have decided thi s fo r American commanders, witho ut any American representati ve hav ing a vote. Under the new arrangeme nts our lia ison officers at B.P. wi ll be in a position to insist that "E" information that is sent. fo r example. to General Montgome ry is also sent to General Clark, if it might have a bearing on hi s ope rations; and if the decision is adverse they can take the maile r up th roug h the appro priate chan nels. 38. We ag reed with the Briti sh to keep Lt. Col. Taylo r's staff fo r the "E" ope ration at B.P. to the m inimu m necessary to do the job; and the Brit ish have ag reed wit h us that the minimum is 4 first-class officers at B.P. We now have one offi cer on d uty, and Lt. Col. Tay lor is te mporarily spending most of his time at the same

SIGNALS INTELLIGENCE

130

work. One other officer has been assigned to the job, and will go as soon as orders and transportation can be obtained for him. Efforts are being made to find 2 others, properly qualified for the job. The liaison orficer at the War Office has already been selected and will start working in the near future. 39. Factors bearing on the detemlination of what reserves of manpower should be built up for the job at B.P. are (a) the desirability of rotating the officers and (b) that, when and if the need arises for American forces on a cross-channel or other operation to have special security officers to handle "E" intelligence for the commanders, they wi ll have to be men trained at B.P. and the Special Branch officers will be likely candidates.

**• From SRH 005. Excerpts ii'om the Military IlIlelligence Service's European

Summary of April 11, 1945 demonstrate the quality of illlelligence about German capabilities alld illlel1liollS that could be derived from ULTRA. EUROPEAN SUMMARY OF II APRIL 1945 No.45- 101 By Auth.

Copy No._

NOTE: No one, without express permission from the proper authorities, may disseminate the information reported in this Summary or communi cate it to any other person. Those authorized to disseminate such information must employ only the most sec ure means, must take every precaution to avoid compromising the source, and must limit dissemination to the minimum number of secure and responsible persons who need the information in order to discharge their duties. No action is to be taken on infonnation herein reported, regardless of temporary advantage, if such action might have the effect of revealing the existence of the source to the enemy. The enemy knows that we attempt to exploit these sources. He does not know. and must not be permitted to leam, either the degree of our success or the particular sources with which we have been successful. I. Western Front: a. Reorgalli:atioll of Armed Forces chain of cammalld: An order issued by Hitler late on 7 April made sweeping changes in the organization of the chain of command in the West. Operational contro l of the Front is now divided between Cin-C West and "C-in-C Northwest," command of the forces in the Ruhr pocket is separated under Army Group " B," and those HQs, plus Naval Chief Command West, are subordinated directly to the OKW. In addition. a new command. "C-in-C

PRODUC ING INTELLIGENCE

13 1

Nethe rlands." has been setup in preparation for "the defe nse of Fortress Holland."' Hitler 's order prescribed the boundary between C-in-Cs West and Northwest as a line running eastward from Paderbom to Schoenebeck (7 m. SE of Magde burg). Details of the reo rgani zation directed by thc order are in substance as foll ows : (I) Subordillated directly 10 the OKW (a) "C-in-C Northwest (A rm y Group ' H T (b) Arm y Group "B" (c) C-in-C West (d) Naval Chief Command West (2) Subordilloted to C -ill-C North",est: (a) 'The f0ll11er sphe re of command of Ope rations Staff North Coast," the staff of which is made available to C-in-C Northwest. (b) "C-in-C Netherlands. to whom the Twe nt y-fifth Army. Commander Armed Forces Netherlands. Admiral Netherlands. and GAF forces in Holland are subordinated. " C-in-C Netherlands "is pe rsonall y responsible to Hitler for the defe nse of Fortress Holland; as soon as land communications to the Reich are c u\. the (previou s ly announced) comprehe ns ive reg ulation s regardin g Fortresses will gove rn hi s relations wi th the Reich Commissar for Occupied Areas of the Netherlands." (c) Amleegruppe Student (d) First Parac hute Amly (e) Wehrkre is XI (3) S ubordinated to C-i n-C West: (a) Aml Y Group "G" (wi th the First and Seventh Armies) (b) Eleventh Army (c) Nineteelllh AIl11y (d) Wehrkreise V, VII, IX , and XIII. (4) Subordinated to Army Group "B": (a) Fifth Panze r Army (b) "Fifteenth Army (A rmee Abt. vo n Luell w it z)" (c) A ll other unit s and personne l o f a ll branches of the Armed Forces in the Army Group area (Ruhr pocket). (5) GAF tactical forces: The organization of GAF tactical forces employed in the West is to conform to the organization of the fi e ld army. A GAF command staff is to be linked w ith each of the gro und commands (presumabl y Luftflolle Re ich in the north and GAF Command West in the south ... ). In add iti on, all flying and Flak uni ts su pporting the field army are to be subordinated operati onally 10 C-in-Cs Northwest and West respective ly. GAF support of Army Group "B" will be arranged speciall y. b. Location of HQs fo ll ow ing the command reorgani za ti on:

( I) C-in-C West moves southeast: On 28 March the HQ of C-i n-C West was in a tunnel at Milseburg. 10 m. E of Fulda . . . .The next da y Genobst. Jod I.

132

SIGNALS INTELLIGENCE

Chief of the OK W Operations Staff, stated that the area west of Ohrdruf (9 m. S of Gotha) was envisaged as an HQ location for C-in-C West " in case of need" .... On 10 April , the following message was sent by an unidentified authority : "As of 0800 hrs on the 10th, the temporary battle HQ of C-in-C West is in Hirschau (40 m. E of Nuernberg and about 100 m. SSE of Ohrdru!). The fin al battle HQ (special train D) will be reported later. "

*** c. Possible lIew commallders: Hitler's order of the 7th did not identify the incumbents of the new ly created posts, but slight suggestions are afforded by the language of the order and subseq uent messages. Although Blaskowitz has been Cin-C Army Group " H" since February, the mo ve of hi s HQ to western Holland seem s incon sistent with hi s assumption of the command into which that Army Group has been merged, and is rather an indication that he may be the new "C-i n-C Netherlands. " Genfidm. Busch, who has been CG Operations Staff North Coast, is a like ly incumbent for "C- in-C Northwest." Hitler 's order specified that thi s high command was to have subordinated to it "the former sphere of command of Operations Staff North Coast," and o n 8 April C- in-C West addressed to "C-in-C Northwest, via Operations Staff North Coast" an inquiry as to when command over Army Group " H," including Armeegruppe Student and Wehrkreis XI , "can be taken over." It is assumed that Genfldm. Kesselring continues as C-in-C West, Genfidm. Model as C-in-C Army Group " B" and Admira l Krancke as Naval Chief Command West. d. Har: sector: ( I) Appearallce of a new Army: On 9 April C-in-C West stated : " Hitler has ordered the Harz (Massi!) as the asse mbl y area for the Twelfth Army. The foremost division of the Twelfth Army is the 85th Inf Div . .. . (portion missing) . ... Any interve ntion by C-in-C West, the Eleventh Anny, or other staffs, is expressly forbidden .... " Note : The new Army is to assemble in the region which Wehrkreise IX and XI have been ordered to prepare for defense .... On 2 April Hitler stated that six divi sions "being set up" were to assemble in the area Hannover - Northeim - Brunswick - which is just northeast of the Harz and which is now being overrun by the Allies. That order said that " the later tasks of this ArnlY will be ordered at the proper time" .... On 26 March the cadre personnel of the 85 th Inf Div, which is being set lip anew inside Germany, were ordered to move from a troop training g round north of Wuerzburg to one just west of Berlin. (2) Ull its to be refilled 0 11 the edge of the bailie :olle: (a) Also on 9 April, C-in-C West subordinated the 394th Assault

PRODUCING INTELLIGENCE

133

Gun Bde, which was assembling just north of the Harz Massif, to the Eleventh Army. The order stated that the Bde was to be rested and refitted, and that guns were to be provided as quickly as possible. Until its "operational capacity" was restored, the Bde was, "within the scope of the Harz defense, to take over defensive and blocking tasks in its rest and refitting area." (b) On the same day (the 9th) C-in-C West ordered that a battle group of the 9th pz Div to be transferre d to the area of Sangerhau se n (20 m. E of Nordhausen) for " further rest and refitting." While there, the battle gro up was also to " prepare for defense, take part in constructing blocking positions, and operate if the Allies attack." (3) Kesse/ring order to the Eleventh Army on 7 April: "( i) In order to gain time, the Eleventh Anny, using no more forces than are necessary, must fight a battle of bitter delaying resistance on its western front , the rear of which rests on the Harz Massif and the mountainous country to the west. Thi s gain of time will serve to make the strong sectors behind the Army 's front capable of defense and to develop the Harz Massif into a blocking Schwerpunkt of the first order. "( ii) It is of utmost importance for the Army, with all speed, to shift to its southern wing the forces which are to be economized on its front facing west. Then, driving toward Langel salza and Gotha (respectively 10 and 20 m. SE of Mue hlhausen), those forces can fall on the Allied flank with adequate striking powe r. "( iii) At the same time it is most important to take immediate countermeasures to forestall the poss ib il ity of the Army 's weste rn front being rolled up from the north and south. Thi s means that , on the north, all roads and tracks in the area Gronau - Alfeld and to the west (i.e., an area about 20 m. S of Hannover) must be rende red impassable with adequate defense; and, on the south , the Allies, as they feel their way northward from Mue hlhausen, must be attacked on the flank s. Here the Eleventh Arnly 's conduct of operations must be adjusted to the utmost mobility and flex ibility. "( iv) The area Nordhausen - Sonderhausen- Kelbra - Schmuecke De pression « i.e., east and northeast of Muehlhausen» must be turned into a firm defensive block. Farther west, local strongpoints such as Uslar, Hardegse n, Northeim , Adelebsen, Harste, Goett ingen. Dransfeld and Witzenhausen «all northeast and east of Kasse l» must be so strongly developed and adequately manned that they will hold firnl under efficient Officers Commanding Local Ope rations. Because of the Allies' well known by-passing tac tics, the flank s of these places must be covered and adequately protected. Covering force s in effective strength , with patrols out to the sides to prevent surpri ses, must be in position at least I 1/2 kilome te rs in front of blocking sectors. The All ies must be brought to a dead stop in this system of blocks."

134

SIGNALS INTELLIGENCE

Note: By 10 April Allied troops had overrun the Gronau - AlfeJd area and the local strongpoints above Kassel , thus rolling up the western front of the Eleventh Anny from both the north and the south.

*** For the Assislant Chief of Staff, G-2 : CARTER W. CLARKE Brigadier General, GSC Deputy Chief, MIS

Distrihlllioll: Within the War Department Secretary of War Chief of Staff CG,AAF Deputy C of S Deput y Commander, AAF AC of S, G-2 AC ofS. OPD AC/AS, Inte llige nce AC/AS. Plans Deputy AC of S, G-2 Chief, MIS

Outside the War Department The President C of S to the Commander in Chief C-in-C, U.S. F leet DDMI, BAS

CHAPTER VII Disseminating the Product The twin breakthroughs achieved by the Army in the field of signal s intelli gence in mid - I 943- making an entry into Japanese military traffic and reaching an agreement with Bletchley Park to share the Briti sh COMINT take from Gennan sources- made it possible for the first time to provide American field commanders with immediate intelligence of operational use. This posed a major problem. How could thi s ultra-sensi tive intelligence be di sseminated in a timely fashion without ri sking compromise of the source which would undo the whole effort. The U.S. Army came up with the so lution of adopting British sec urity regulations and creating a Special Security Officer system under Special Branch to handle distribution of signal s intelligence to the field . Specially picked and trai ned officers, operating under War Department control and utilizing a special communications system, would serve as the conduits for high -g rade COMTNTULTRA , as it was now called. These officers would act both as secu rity watchdogs and as guides to properly interpreting the material. In SRH 026, Marshall Leller to Eisenhower, the Anny Chief of Staff explain s the new security system to the Commander, European Theater of Operations. SRH 033 , History of Operations of the Special Security Officers , provides a general overview. Excerpts from SRH 031 , Trip Reports - ULTRA in the Mediterranean; SRH 022 , ULTRA and the Seventh U.S. Army; and SRH 032, ULTRA in the Southwest Pacific Area, contain reports from individual Special Security Officers on condi tion s in the field.

136

SIGNALS INTELLIGENCE

From SRH 026. As a resull of Ihe agreemenl iJelWeen Ihe Brilish and Ih e War Departmenl 10 share cryplOlogic il1lelligellce, "VIII'a" malerial could iJe sel1l 10 American field commanders tinder slricI seCllrity provisions. In Ihis lellel: Army Chief of Slaff George C. Marshall explains Ih e arrangemenls 10 his principal suiJordinale in Europe, General Dwigfu D . Eisenhowel: The securiIy regulations of April 1944 10 which Marshall refers illlroduced Ih e TOP SECRET classification to the V. S. Army for Ihe firsl lime.

WAR DEPARTMENT THE CHIEF OF STAFF Washington, D. C. March 15, 1944 Dear Eisenhower: You are undoubtedly aware of th e s upreme imporlance which the War Department a!laches 10 intelligence known as "U ltra." This intelligence is secured by the British from reading German enciphered radio communications. The a!lached Tab sets forth the basis upon which German "U ltra" intelligence is made available to American field commands. Please give thi s matter your personal a!lention, and take all necessary steps to insure that the security regu lations governing the dissemination of " Ultra" intelligence are meticulously observed. The arrangements described in the attached Tab are to be fully carried out. Faithfully yours, /s/ G.c. MARSHALL General D.D. Eisenhower, Supreme Headquarters, Allied Expeditionary Force, London, England. I. A large volume of highly important military and air intelligence is derived by the British from reading Gennan radio communications enciphered in high-level German c ipher systems. Thi s type of intelligence is known as '"Ultra" intelligence. 2. The basis upon which Gernlan Ultra intelligence is made available by the British to the US War Department and to American Field Commands in active theaters is gove rned by an agreement mad e in June 1943 between the US War Department and the responsible Briti sh organization. 3 . The British organization which produces Ultra intelligence is th e Government Code and Cipher School , the Director General of which is Bri gad ier Sir Stewart Menzies.

DISSEMINATING THE PRODUCT

137

4. Administration of the agreement of June 1943 is controlled by the British [text withheld] through Brigadier Menzies, and by the US War Department through the A. C. of S., G-2. 5. To carry out the agreement of June 1943. and to assist in and supervise the dissemination of Ultra intelligence to American Commands in the European Theater, a detachment of MID, the title of which is MID , War Department , London, has been detailed for service in the European Theater. This detachment is under the direct control of the A. C. of S. , G-2, and the Commanding Officer of the detachment is the representative of the A. C. of S. , G-2, on all malters relating to or affecting Ultra intelligence in the European Theater. 6. Security regulations governing the di ssemination and handling of Ultra intelligence within the European and Mediterranean Theaters have been approved by the US War Department [text withheld] and will become effective on I April 1944. 7. Under the agreement of June 1943 and the security regulations of I April 1944, the basis of disseminating Ultra intelligence to American Commands in the European Theater is as follows: lh Ultra intelligence produced by G.c. & C.S. (including the component of the US Signal Corps at G.c. & C.S.) is disseminated to Field Commands only by means of special communications channels established and controlled by the Director General, G.c. & C.S. Q., Communications between Field Commands which discuss or mention Ultra intelligence must be passed only over the above special communication channels. ~ American officers participate at G.c. & C.S. in the selection of Ultra intelligence for dissemination to the Field Commands and in the preparation of the messages in which the Ultra intelli gence is so disseminated. These officers are attached to MID. War Department. London. d. The receipt and distribution, at the Field Commands, of messages containing or relating to Ultra intelligence, is handled by Special Liai son Units fumished and controlled by the Director General, G.c. & C.S. The personnel of these units includes American officers attached to MID. War Department. London. lh The Special Liaison Units distribute Ultra inte lligence messages only to officers at the Field Commands who are li sted as entitled to receive Ultra intelligence. The elig ibilit y of officers for li sting as authorized Ultra intelli gence recipients is governed by the securit y regulations of I April 1944. The Director General, G.C. & C.S., and the Commanding Officer of MID, War Department, London, administer the provisions of the agreement relating to admission of American officers in your theater to the list. L One or more American officers assigned to MID. War Department. London. will be detailed to each American Field Command, which receives

138

SIGNALS INTELLIGENCE

Ultra intelligence. These officers will be subject 10 the administration and discipline of the Command to which they are detailed. They will work under the control of the G-2 or A-2 of the Command as part of hi s staff. They will have had a period of training at G.c. & C.S., and. if possible, with operational commands in the Mediterranean Thearer, and this training will be directed IOward equipping them to use Ultra intelligence effectively and securely. Their primary responsibility will be 10 evaluate Ultra intelligence, present it in useable form to the Commanding Officer and to such of hi s senior staff officers as are authorized Ultra recipients, assi st in fu sing Ultra intelligence with intelligence derived from other sources, and give advice in connection with making operational use of Ultra intelligence in such fashion thar the security of the source is not endangered. If at any time the flow of Ultra intelligence is not sufficient to occupy fully the time of these officers. they may be used for other related intelligence assignments. 8. The Commanding Officer and the G-2 's and A-2 's of all American Field Commands in your theater, which receive Ultra intelligence, will take all steps necessary to insure that the requirements of this leHer and of the security regulations of I April 1944. are fully carried oul. In order to sa feguard the continued availability of this enormous ly important source of intelli gence, it is vital that these security regulations be meticulously observed. and thar all personnel entitled to handle or receive Ultra intellige nce take all possible precautions in connection with its handling and use. When operational action is take n on the basis of Ultra intelligence, the utmost care must be taken, by means of proper cover, to insure that the action does not reveal or in any way suggest that this source of intelligence is at our disposal. 9. h is of particular importance that: !l.Ultra intelligence be transmilled 10 the Field Commands only by those spec ial channels mentioned in paragraphs 7a and b above, and that communications discussing or mentioning Ultra inte lligence be transmilled only by those c hannels. Q.Uhra intelligence be discussed orally only with personnel of the field commands who are li sted as authorized recipients of Uhra intelligence; and £. Full facilities and opportunity be extended to the MID. War Departme nl. London, officers detailed 10 Field Commands in order 10 enable the m to perform their duties full y a nd effectively. The Commanding Officer, MID, War Department, London , and hi s prin c ipa l assistants, will visit th e Field Commands as occasion requires to consult with the G-2's or A-2 's on methods of handling or using Ultra intelligence and on the scope and method of servicing Ultra intelligence from G.C. & C.S. to the field commands. 10. The contents of thi s le Her will be co mmuni cated to the Commanding Officers and G-2's or A-2's of all American Field Commands in yo ur theater which receive Ultra intelligence.

DISSEMI NATING TH E PRODUCT

139

From SRH 0 33 . This docllment was prepared ill 1945 as an introdu ctioll to th e num ero us individual acti Fity reports prepa red by Special Security Officers and Special Security Representati ves assigned to the Fariolls theaters. Selected reports are reprodu ced later in this chapter. HI STORY OF TH E OPER ATION S OF SPECI AL SECU RITY O FFICERS ATTAC HED TO FlELD COMM AN DS 1943 - 1945 I. The Special Securit y Officer (SSO ) system was organi zed for the purpose of providing a means of rapid and secure di ssemination o f Ultra intelligence to operatin g comm ands . Th e sys te m was deve lo ped and put into ope rati on by the o ld Special Branc h, MID. At the time of the June 1944 reorgani zati on of the Military Intelligence Se rvice the fun ctio ns o f admini ste ring the SSO system we re tran sfe rred to the new Special Branch, although certain o f the dissemination functions were retained by the Director of Intelli gence, MI S. Moreover, all policy problems were retained unde r the direct control o f the De puty Chief, MIS. Thi s divi sion o f a uthorit y did not prove who lly satisfactory. In Fe bruary 1945 complete responsibility for the admini stration of the SSO syste m was vested in the o ffi ce of the Deput y Chief, MIS , where it still remains. 2. Thi s hi story is concerned with the developme nt of the SS O system as a whole and will cover the e ntire pe riod of its ope ration regardless o f the pl ace in the MIS organi zation whe re the functions may ha ve been located. Because each theate r o f ope ration presented di ffe rent problems, it was necessary to place a large measure of responsibility in the senior SSO in each theate r (eventually his title was changed to Special Security Representative, abbre viated as SSR) for the supply of Ultra to commands operating in hi s area. A history of the ope rations within each theate r is there fore o f much g reate r value in alle mpting to evaluate the me rits of the system. For this reason detailed histories by theate r have bee n prepared and are attac hed he reto. The purpose of this introducto ry stateme nt is to descri be the ste ps whi ch led up to the organization of the system and the manne r in which the '"home office" was operated in support o f the separate theate r organi zations. 3. The need fo r the ficld di ssemination of Ultra did not become ac ute until 1943. Arlington Hall did not make it s first entry into mai nline Japanese milita ry systems unti l Marc h o f that yea r, and it was no t unt il the fall o f 1943 that an y wo rth while inte lligence of an operati onal nature was being produced from these military syste ms. Hi gh level diplomati c traffi c had been available since 194 1 but its va lue was strategic in nature and ve ry seldom did a nything o f immediate tacti ca l impo rtance deve lop from thi s source . The re we re in ex istence during the pe riod from 194 1 to 1943 certain special c hannels of communication by means of which Ultra inte lli gence could be made avail able to va rious field commands. The Navy

140

SIGNALS INTELLIGENCE

had a direct Ultra channel to CINCPAC at Pearl Harbor, with an extension to the Commander, Seventh Fleet in Bri sbane. Genera l MacArthur had his own cryptanalytic organization , Central Bureau Brisbane, which had special channels of communication to Signal Security Agency (Arlington Hall) [text wi thheld] There was no direct intelli gence channe l for Ultra from the Military Intelligence Service to any of these commands. 4. German military Ultra in great volume and of the hi ghest operational significance was being produced by GCCS in London. The results we re disseminated over a spec ial Ultra intelligence channe l maintained by Briti sh Special Liai son Units (known as the SLU organization). Throughout the North African campaign the SLU organization served key American officers with thi s inte lligence in accord with British securit y practices. The Military Intelligence Service did not participate in this serv ice, and as a matter of fact , was not completely aware of its existence until the summer of 1943. Gennan military Ultra was never produced by Arlington Hall, the primary reason being that it was not possible to intercept German military traffi c on the North American Continent. 5. The first steps taken to unify the method of fi eld di ssemination of Ultra and to insure that all Ultra intelligence was made available to the appropriate operational commands, was to come to an understanding with the British. In the spring of 1943 an agreement was entered into between G-2 and [text withheld) London [text withheld] providing for complete cooperation in the Ultra intelligence fie ld. By the terms of this ag reement, the Briti sh assumed primary responsibility for the production and ex ploitation of German military Ultra, and the United States Army agreed to devote its attention to the production and exploitation of Japanese military Ultra. Full exchange of the products of each service was provided , and the principle was established that each service, in cooperati on with the other, wo uld establish methods of dissemination to its own operational commands. [tex t wi thhe ld] The subject need not to be further discussed here other than to say that the Briti sh SLU organization [text wi thheld] was the pattern upon which the SSO system was developed for dissemination of Ultra to Pacifi c commands. 6. The German problem having been solved in principle, attention was then directed to providing a system for the Pacific. The peculiar nature of Ultra intelligence required a system of dissemination which, in some respects, was contrary to establi shed principles of command responsibility. The high level Japanese military cryptographic systems were of such difficulty that onl y a tremendous organi zation equipped wi th ex pensive and bulky electrical tabulating machinery could prod uce any results. Consequentl y, detailed items of immediate tac tical significance to an Army commander in the fi eld could be read at Arlington Hall , but could not be produced by an agency located in the theater and near the Army commander interested . Therefore "combat intelligence" was being produced in Washington, some 15,000 miles away from the front lines. To forward thi s intelligence through normal command channels of communication and through every internlediate G-2

DI SSEMI NATING TH E PRODUCT

14 1

betwee n the War De partment General Staff and the combat commander concenled would de lay its receipt and, in many cases, the particular message would lose all value by the time the commande r finally received it. Mo re impo rt ant were the securit y considerations involved. Experie nce in World War I and the experie nce of the British in World War II had proved that wide dissemination of Ultra very ofte n led to a compromise and complete loss of the source. The bus iness of crack ing codes is mov ie thrille r stuff and the natural reaction of an uninit iated person is to make some comment about it to other responsible officers who would nonna ll y be supposed to be in on the secret. Care less talk by one unin fo nned person to ano the r soon res ults in the "secret" be ing public in fo nllation . To avoid these dangers, the British had adopted several rules of dissemination which had proved highl y successful in maintaining the necessary security. They were: a. Ultra was never to be made available to anyone, from a commandi ng gene ral down to the code clerk who se nt the message, without fi rs t explaining to the recipie nt the securit y questions involved, and how even the most innocent comment might lead to loss of the e ntire source; b. Ultra would be di sseminated onl y via special communication channels which would be used solely fo r lhat purpose. This wo uld prevent a casual know ledge o f Ultra o n the part o f pe rsonne l employed in communication cente rs, and would also ena ble complete control of d issemination within the theater of any messages transmitted over thi s c hannel; c. The special dissemination channels wo uld be maintained by officers and e nli sted men from a central o rgani zati on and would be furni shed onl y to commands which had ac tual need of Ultra inte lli gence. The loca l commander wo uld therefore not have the authori ty to over-rule this personnel on any matte rs involving security. d. Strict rules fo r the di ssemination within the command, limiting the number of officers authorized to see the intelligence, and the uses which could be made of it , would be presc ri bed by the central organization . 7. The Pac ific field di ssemination syste m was dev ised in the summe r of 1943 on the basis o f the principles outlined a bove. It was dec ided to designate the o ffi cers who wo uld maintain the special channels as Spec ial Security Offi cers (SSO 's). SSO's we re to be assigned to the Military Inte lligence Se rvice and remain under the ope rational control of the A. C. of S. , G-2, WDG S. They wo uld be attac hed to the theate r commander for purposes of adm ini stration and d isci pline and wo uld disseminate Ultra in accord with rules announced in security regul ations issued by the War De partment. Authorit y was obtained from the Chief of Staff in the summe r o f 1943 to consult the theate r commande rs in the Pac ific and to o bt ain their conc urrence with thi s plan. T he commanding generals of the Pacific Ocean Areas and the China-Bunll a- India (C BI) Theate r readil y conc urred, but the Commande r in Chief, South wes t Pac ific Area, raised cert ai n objectio ns to the proposa l on the gro und that the SSO should be ass igned o utrig ht to hi s comma nd ra the r th a n

142

SIGNALS [NTELLIGENCE

remain under the control of the War Department. Col. Carter W. Clarke, then Chief of Special Branch, was sent to Brisbane to confer personally with CINC, SWPA on this difference of opinion. As a result of this conference, General MacArthur concurred in the original plan . 8. The first War Department security regulations goveming overseas dissemination of signal intelligence were issued in October 1943. Following the British practice, two classifications were established: ULTRA DEXTER and DEXTER. Subsequently these terms were increased to three, as follows : ULTRA [text withheidi PEARL [text withheld] THUMB [text withheld] .... The regulations goveming Pearl and Thumb were subsequently revised, in conjunction with the Navy, and the single code word PINUP was adopted for all low level signal intelligence. 9. The original Ultra Dexter regulations were imp lemented in the fall of 1943, when special security officers were sent to the following theater headquarters: Southwest Pacific Area, Brisbane (Major James Ashby, Jr.); China-Bunna-India, New Delhi (Captain John FB. Runnalls); Pacific Ocean Area, Fort Shafter, T.H. (Major Edwin E. Huddleson, Jr.). The original plan was to limit overseas dissemination of Ultra to theater headquarters and to await recommendations from the spe-

cial security officers before making provision for broader dissemination. It will be recalled that the operational situation in all three theaters in late 1943 and the first half of 1944 was not sllch as to require dissemination to subordinate operational commands. Accordingly, all three special security officers undertook to serve the respective theater commanders and their staffs. To accomplish this each officer carried with him his own set of cryptographic equipment (SIGABA) for the enciphering and deciphering of Ultra messages . Special arrangements were set up to provide pouch service as well, and such material as the Diplomatic Summary was sent regularly by special Top Secret pouch. It should be noted that both radio and pouch communications were sent to the special security officer personally. Thus he received radio communications in enciphered form and did the actual deciphering himself. He was therefore the only person to see sllch messages in the clear, and showed it personally to the theater commander and other specified staff officers. At no point did the contents leave the control of the special security officer. POlich material was delivered to him unopened, and he thereupon showed relevant material to appropriate headquarters officers, retaining actual custody of it himself. This system proved both secure and efficient. Except for certain considerations peculiar to SWPA [text withheld] the special security system won the approval and confidence of theater officers concerned. 10. Meanwhile, preparations were underway for the establishment of a special security system to serve in ETO in connection with the expected invasion of the spring of 1944. A larger number of officers were recruited and sent to the ETO for training . ... Accordingly there arose the need of a Table of Organization which would provide vacancies for enough special security officers to fill the needs of both the European and Japanese wars. In December 1943 a staff study was pre-

DISSEMINATING THE PRODUCT

143

pared which described the special security requirements and stated the estimated personnel needs. Approval was thereupon given by the Adjutant General for a field quota of 80 officers. In July 1944 thi s quota was enlarged to 172 officers and 65 enlisted men. a total which proved sufficient for the duration of the war. 11. In the spring of 1944 sufficient information had become available, based on recommendations of special securit y officers and upon projected operational plans insofar as they were available to MIS , to pernlit a dec ision as to the extent to which Ultra should be disseminated below the leve l of theat er headquart ers. Accordingly, the regulations were rewritten (10 July 1944) to allow dissemination to the level of Army (or equivalent Air Force Formations) and to Corps leve l when the Corps was operating independently. In April and MayoI' 1944, by way of preparation for the revised Ultra program, officers were recruited for SSO duty in the Pacific and Far East. The first group, numbering about 20, began a training period on 3 July 1944 and were sent overseas about I September 1944. This group included a full complement for CBI and a group of officers for SWPA . The training course consisted of a week's tour of ASA, two weeks' training in the specific duties of the special security officer (study of regulation s; lectures by returning officers familiar with the theater concerned; consultation with relevant sections of OPD; cryptographic security; communications, etc.) and from four to six weeks straight intelligence training, which included the Far Eastern Intelligence Course and selected reading of Ultra material s; where possible the officers spent from one to four weeks actually working on Ultra traffic as members of the various Ultra sections of MIS. The value of this latter aspect of training cannot be over emphasized. This first group which went overseas have all testified to the importance of an adequate intelligence background. By and large, this g roup was well trained, but as the operational situation grew morc active it became necessary to send additional special security officers on very short notice. In many cases such officers had to be dispatched with very little training or background in Ultra. If there were one lesson to be learned from experience in connect ion with the SSO program in the Far East, it is thi s: a large potential pool of security officers should be recruited earlyas early as possible and long before they will actually be needed overseas. They should function as working members of MIS- thi s is the very best type of training- and be available for overseas assignment as the need arises. The greatest single handicap in the past administration of the SSO system was the lack of adequately trained personnel to meet the suddenly increasing needs of the Pacific war. While it can be said that the program was carried out efficient ly in the face of numerous obstacles. and that the sudden expansion in SSO requirements could not have been anticipated. the ultimate results could have been more satisfactory if we had had too many qualified personnel available rather than too few. Admitting that it is always difficult to obtain as many highly qualified officers as one would like. this phase of the problem should be given particular and immed iate attent ion in the event that the need for an SSO organization should ever arise again.

144

SIGNALS INTELLIGENCE

12. The manner in which the SSO program was carried out in the respective theaters is fully explained in the individual histories. It remains only to discuss brieny the functioning of the home office. Before the G-2 reorganization of June 1944, the administration of the overseas dissemination was handled by two officers in Special Branch. After the reorganization, when the SSO field quota expanded from 3 to some 65 officers in the Pacific and Far East, the administration was made the responsibility of the newly const ituted Special Branch. A total of three officers were actually engaged in the admini strative phase of the SSO program. These officers were concerned with the fOllowing aspects of the SSO program: recruiting and training of SSO's; review and revision of Ultra, Pearl , and Thumb regulations; handling of all administrative correspondence, by radio and pouch, between MIS and SSO's overseas; liaison with interested sections of the intelligence group in connection with intelligence requests from SSO's; coordination, where necessary, with ASA, OPD, OCS, the Navy, and wi th various other organizations of the War Department; all personnel problems, including promotions, travel orders, securi ty clearances for both officers and enlisted men; a myriad of miscellaneous matters which cannot be conven iently cata logued but which arise in any unit of the size and scope of the SSO organization . While it is a matter of record that there was no serious breach of security throughout the war, and while theater officers have uniformly testified to the quality of the serv ice they received , it is subm itted that a larger home office, more fully staFfed both with officers and civ ilians, wou ld be desirable " next time,"

13. In January 1945 the functions of Special Branch were transferred to the Office of the Ch ief, MIS and placed under the direct supervision of the Deputy Chief. Inasmuch as the Special Branch program was under the direction of the Deputy Chief from the outset, this transfer proved more satisfactory, since it eliminated an extra step in the chain of command and simplified the matter of coordination with the Director of Intelligence, a major part of the Special Branch job. Again the duties were performed by three officers, although this number was increased to five during the summer of 1945. There were no substantial changes in procedures as a result of the above mentioned transfer. 14. Apart from personnel , the most important single factor in the eFficient and sec ure operation of the SSO system, was the special communications channel, which worked magnificently and proved eminently satisfactory throughout the war.

DISSEMINATLNG THE PRODUCT

145

From SRH 03l . In preparation for a future assignment as a Special Security

Officel; all Air Corps lieutenant colonel went to the Mediterranean Theater of Operations to observe how the British had organized the distribution of Ultra material in that area. It should be remembered tilat Special Brallch serviced air commanders as well as the groul"/{1 Army. MAAF refers to Mediterranean Allied Air Forc es; in today's military parlance, it was a combined service operation rather than a joint aile. 24 June 1944 SUBJECT: Report on Mediterranean Trip. TO: Commanding Officer, MIS, WD, London. FROM: Lt. Col. Leslie L. Rood, A.C. I. This is a report on my visit to the Mediterranean Theater of Operations in June 1944, to observe the use of air specia l intelligence. Because the London office is already familiar with the genera l functioning of the Mediterranean organization this report is on ly of a supp lementary nature. 2. Mediterranean Allied Air Forces, a joint British-American Command, is the supreme air headquarters of the Mediterranean Theater. It has a rear headquarters at Algiers and an advanced headquarters at Caserta. Under it are Strategic Air Force Tactical Air Force, and Coastal Command. The Strategic Air Force is an almos t wholly American Unit, the 15th Air Force, operating from Bari. They have one British group of Wellingtons for night bombing. The 15th is an operational command in that it plans and directs missions. Tactical Air Force at Caserta, on the other hand , is a British-American Command that is not operational. It has under it (I) the 12th Tactical Air Command (American) which gives close support (fighter bombers) to the 5th Anny, (2) Desert Air Force (British) which gives close support to the 8th ArnlY, and (3) two wings, one of medium bombers and one of fighters. which operate at a greater distance behind the enemy lines. Coastal Command, with defensive, convoy and coastal missions. is spread out over the whole Mediterranean. Tactical and Strategic are in Italy. 3. I departed from London on the evening of 26 May 1944 by sleeping car and arrived in Newquay, Cornwall, the next morning. I should have taken off from St. Mawgans that night , but due to engine trouble did not get off until the fourth night. The plane arrived at Casablanca the next morning and by evening I was in Algiers. I spent three days with MAAF (rear) and took off for Caserta on 4 June. For the succeeding three days I was there with MAAF (adv.) and TAF. On 8 June I went to Bari by courier plane to spend three days with the 15th Air Force and 276th Wing. Returning to Caserta on the II th I was unable to go forward to 12th

146

SIGNALS INTELLIG ENCE

TAC at Rome until the 14th because they were disorganized by the rapid advance. When I did reach Rome I was able to stay with them for only two days because they again split into three echelons and moved on. On the evening of the 17th I left Rome; on the morning of the 18th I left Caserta; the nex t morning I departed from Algiers; after a delay of one day in Casablanca I took off for the U.K. and was in London on the evening of 21 June 1944. 4. The two MAAF special intelli gence sections operate like miniature Air Ministries. At each headquarters the four or five officers processing the material are segregated in an out-of-the-way room where they do their work. Each officer has a specialty such as order of battle, " Y," supplies, Balkans, or targets and as the expert on that subject, he wri tes part of the daily digest which is prepared for the generals. The senior officer of the group usually carries the di gest around, so that he can answer any questions which arise. They write very few appreciations, nor do they have any direct connections with operations. Their prime purpose is to be completely familiar with the material so that they can interpret it to their operational people. Secondarily, they compile a weekly order of battle which is publi shed at several levels. MAAF (rear) al so sends out a week ly suppl y summary at the highest level. 5. The 15th Air Force at Bari impressed me as a very smoothly fun ctioning organization and it s "A-2 Liaison Officer," Major Murphy, as an able part of it. Each morning the staff meets with the A-3 (acting as the operational chief of staff) to plan the mi ssions for the subsequent day. First the weather officer delimits the possible areas of operation by his forecast; the A-2 names the targets having priori ty in that area; the damage analysis officer advises the A-3 to what ex tent the suggested targets have already been destroyed. Attention then being officially centered on several targets (the staff had already infornlally decided on these and spread the proper target charts before the meeting) the flak officer is called upon for the number of minutes the ship s will be in flak if approaches are made from ce rtain angles. Major Murphy is then called upon for an est imate of whe re, how man y, and what kind of e nemy pla nes will be encountered. Along with the weather officer he stakes hi s reputation dail y since both of their pred ictions are checked by the next evemng. After quite a success ion of direct questions by the A-3 and unequivoca l answers by the staff, the A-3 fluently dictates a complete operations order. Each officer copies down the part pertaining to hi s responsibility and during Ihe early afternoon submi ts the written annex to the finished order whi ch is placed before the Commanding General for his approva l. Because of hi s daily forecast, Major Murphy is known to most people as the order of battle expert but even more important is his servicing of senior officers with spec ial intelligence. He separates fro m the general material that whi ch mighl interest the sen ior officers, annotates it wit h brief notes when necessary, and then

DISSEMINATING THE PRODUCT

147

shows the m the annotated material. He often has to do some verbal explaining and elaborating but he writes no digest. There was a time when special intelligence was important at Bari for its order of battle information, but now when the GAF puts up a maximum of 150 planes to protect the most important targets and none for lesser targets, raids are naturally planned with little regard for the possible opposition. The most valuable material , therefore, is that pertaining to targets, e.g . that a bridge was destroyed, that shipping on the Danube has been stopped or that damage to refineries has caused a gasol ine shortage. Operations people take a very keen interest in anything which gives a clear cut statement of the damage done by their bombers the day before. With the increase in blind bombing this is even more true. Most of the intelligence at the 15th Air Force consists of target work, i.e. photography, collating of agents reports and economic studies of objectives, and bomb damage analysis. There is a large target section staff at MAAF (adv.) which serves the 15th Air Force almost exclusively by preparing target charts and 15th Air Force also has a large target section which selects objectives and analyzes bomb damage. Since no one in the target section at Bari has special intelligence, Murphy helps them with his s uperior wisdom quite frequently. In attempting to s ummarize for me what special intelligence supplied to the whole intelligence organization at Bari , Murphy said that its primary contribution was a "negative influence." Mos t of the information Murphy gets from special intelligence is also procured from photos. PW's , pilots' reports. and " Y," but in them the good is mixed with the bad. Special intelligence tells Murphy and his superiors what not to rely on. Rarely does he get anything of pure operational value, something which causes him to rush into A-3 demanding that a mission be layed on. 6. TAF 's functions are difficult to understand because of its intermediate role below the policy level of MAAF and yet above the target selection level of TAC and DAF. Its role is further complicated by its being an operational headquarters over its medium bomb wing but a non-operational headquarters over its close sup-

port commands. It is primarily a coordinating command. Its greatest interest during my visit was in the progress of its campaign to choke off all supplies to the battle area. Special intelligence provided very good material on this subject. The segregation of special intelligence is not very noticeable here because only the chief intelligence officer, W/C Wiseman, and his assistant, Captain Austin, have it. Their knowledge is used through attendance at all planning conferences, some of which are at the highest level. They show selected material to the senior officers and do not write a digest. 7. Di scovering how Major Corning, at 12th Tactical Air Force, uses s pecial intelligence was not easy. He keeps no records that one can analyze. He is the A-2

148

SIG NALS INTELLIGENC E

and the o nl y pe rson doing pure inte lli gence wo rk so s pecia l inte lligence is not neatl y segregated for the visitor's examinat ion. Coming operates o ut of hi s shirt pocket. me rges all sources in his mind, and is nOl g iven to pat statements about what part of his tota l knowledge is supplied by special inte lli ge nce. 12th TAC, in supporting the 5th A ml Y is meeting abso lute ly no enemy planes so natu rall y order of banle is forgone n. Likew ise, the occasiona l damage re ports on spec ific targets are of doubtful use beca use TAC allac ks vast numbe rs of small targets each day such as motor transport, bridges, rail way c uts and dumps. What is important is that special inte llige nce shows TAC very clearly that its prima ry mi ss ion, c Ulling the lines o f communicati on and suppl y, has been successfull y acco mplished . The G erman in ability to defend has been caused to a great extent by the lack of supplies. In the fas t mov ing situation prevailing d uring my visit spec ia l inte lli ge nce ind ications of whe re the enemy might make a stand was a lso o f gre at value to TAC in decidi ng whe re to establi sh its a irfie lds as it advanced . TAC's inte lligence set up is ve ry simple. The broad pl ans of anac k a re made by Arm y; the target section is run by the A rm y and a ll TAC gets is a set of photos a nd a request for missions; there is no air o rder of bailie; a lmost a ll the spec ial inte ll igence coming through during my visit concerned the AmlY and was onl y of inc identa l inte rest to the A-2 . The A -2 does the pure inte lligence wo rk and hi s e ight subordinates are primaril y engaged in compiling operations summaries, an A-3 job. 8. My visit to the 276th Wing was valuable chie fl y because I saw a comple te " Y" organi zation ope rating under one roof. Thi s organi zation fun ctioning without the complicated refinements of Block F, was " Y for C hildre n. " Spec ia l inte lli ge nce is used at 276 Win g so le ly as a g uiding influe nce. II e nables the seni or o ffi cer to e liminate from the published re port s many of the erron eo u s co nc lu s io n s w hi c h wo uld res ult fr o m d epe n de n ce o n " Y " al o ne. Unfo rtunate ly there is so linie traffic to be picked up in the Medite rranean that the ir reports are comparati vely uni mportant. 9. Seeing fo ur levels of command in the fi e ld immediate ly fo llow ing my visi t to War Station and Air Ministry, the sta ir like decline o f fOflllality in handling spec ia l inte lligence was ve ry noticeable. At one extre me is an e laborate recording and weighing of every deta il and at the othe r is the casua l treatme nt at 12th TAC where the recipie nt sk ims through the pe nc il written pages of the day, tracing some out on the map, re reading some, ignorin g othe rs and in no case writing anything down. At MAA F (rear) quite a bit o f index ing is done so th m records will be available for a ppreciati ons and wee kl y summaries. At MAAF (ad va nced) onl y an orde r o f battle index is ma intained, and at the next lowe r le ve l, TAF, the infomlality approac hes that of TAC in that onl y an occasional note is made. The 15th, like its para lle l command , TA F, uses practicall y no records. 10. It is also noticeable that each level o f command endeavors to stand on it s

DISSEMINATING THE PRODUCT

149

own feet rather than rely on appreciations and digests from superior commands. MAAF (rear) sends little to MAAF (advanced), MAAF (advanced) almost nothing to TAF, TAF nothing to TAC and naturally TAC can have no recipients. What little is sent is treated with not too much respect. Each level serv ices its own generals without attempting to also service its lower echelons. Many of the recipients do see each other quite frequently to exchange views. Probably for thi s reason no great differences of opinion are evident. I I. At War Station and the Air Mini stry, I got the impression that intelligence was 75% special intelligence, 10% "Y," 5% photographs, 5% PW, and 5% mi scellaneous. The figures varied somewhat but were usually definite. The people in the field don't think much in terms of percentages and, if they did, the figures would vary tremendously from place to place because each headquarters looks to intelligence for a different type of information. Sometimes special intelligence can fill the need and sometimes it can't. The usefulness of other sources varies also, depending on what the headquarters is interested in. For commands where few records are kept PW reports mean little because no report is complete within itself. Usually they are skimmed through. Agents reports are seldom read by the A-2 or the order of battle people but they are of some importance to target sections. "Y" in the Mediterranean is sketchy, but along with photo recce of airfields which is fairly good in the Med, does provide good cover for the specia l intelligence order of battle. Probably the most valuab le source in the theater is photography. With the decline in interest of order of battle, targets are of comparatively more importance. The photo tells the pilot where the target is and within 24 hours of the attack photos give fairly complete damage information. Special intelligence does, of course, enable one to exercise great wisdom in reading photos. 12. Covering the use of spec ial intelligence in the field does not present the problems which I understood it did. This is primarily so because receipt of the material does not usually point the way to an operational mission. The material provides background which enables commanders to make the correct decisions: it confirms or negates infom1ation from nom1al sources; it provides basic facts. None of these need cover. 13. At both MAAF headquarters the groups of people processing special intelligence are wholly British. The senior intelligence officers under whom these people work, A/C Woolley and G/C Luard, each have American executive officers, but only Woolley 's executive, Major Tandler, handles special intelligence. Even he does no actual work with it. While in theory this use of British personnel alone to serve a joint BritishAmerican Command may be criticized, there is no practical solution. The Briti sh appear to be doing a good job and their long experience in work where knowledge of the past is invaluable would place any newly assigned American officer in a distinctly subordinate position. In addition they already are overstaffed.

150

SIGNALS INTELLIGENCE

At TAF a lso the chief intelligence officer is Briti sh but thi s provides good bal ance since the remainder of the sta ff is predominantl y A meri can.

DETAILS : I. A he lmet and gas mask are not needed in vis iting Med ite rran ean headquarters. The on ly li se for a helmet is as a wash ba sin at headquarters near th e front and

even there borrowing is ex pectcd. 2. Arrangements for tran sportation take lime and should therefore be initiated

for the next hop as soon as one arrives at a station. 3. Orde rs c ut in London should authorize visits to all c iti es w hic h mu st be reached by ATe. i.e . Al g ie rs and Naples in June 1944. If o rd e rs me ntion on ly Algiers, more orders must be obtained to ge t on to Naples. Transportation beyo nd the ATC te mlinus is by courier plane or airfield hitch hiking for which no o rd e rs are necessary.

4. Cotton uniform s may be pu rchased in Algie rs at the PX.

From SRH 022 . The report of Ih e Special Sec/wily Officer assigned 10 Ih e Sevenlh Army, Ihe American force Ihal rook parr in Ihe invasion of SOI/Ihern France ill AI/gusl 1944. The "SCUISLU link" refers 10 Ih e Brilish-operaled Special Comnl/lniCaliolls UllillSpecial Liaison Ullil Ihal served as Ihe COIIduil for UI,ra inrelligellce produced by Ihe Governmelll Code alld Cypher School al Blelchiey Park.

ULTRA ANDTHE U.S. SEVENTH ARMY AMERICAN EMBASSY OFFICE OF THE MILITARY ATTACHE I. GROSVENOR SQUARE, w.1. LONDON , ENGLAN D 12 Ma y 1945 MEMORANDUM FOR COLONEL TAYLOR Subject: Ultra and U.S. Seventh Army I. Introduction: The following report summari zes the ex pe ri e nce of the Seventh U. S. Arm y in the use of ultra inte lli gence betwcc n 16 August "44 and 8 May '45. The judgments, conclusions, ancl recommedalions are not intended as generali zations necessaril y appl ica ble to e ither air or other ground commands. 2. Ope rations: The SCU/SLU link and ultra spec ialist joined the Seventh Army at Naples, and the station o pened on D plus I , the 16th of August. It had bcen agreed w ith the G-2 that the jo b of the ultra spec iali st was to receive signal s from

DISSEMINATING THE PRODUCT

151

the SLU, process the m. and post the information on a map. For the first few days of operations in South France, the onl y map available was a folding map which was shown periodicall y to the G-2. During thi s pe riod the qualit y of ultra intelli gence was usually high and in the rapidly moving situation was the primary source of tactical information . The G-2 dec ided, the refore. that an uhra war room was desirable, and henceforth the ultra specialist occupied a room in the CP adjacent to the G-2. Also during these early days the volume of uhra was so great that it was soon apparent to the G-2 that he could not personally brief his commander on the content. Accordingly. the responsibility for the briefing of all recipient s devolved upon the uhra speciali st. At no time during operations by the Seventh AmlY was there a formal briefing for all recipients. There can be little doubt that formal briefings are desi rable for they permit a more careful preparation of the material. But the recipients were not amenable to th e idea because

it involved an additional commitment of time in th eir

already busy day. Al so, the demands of the tactical si tuation dictated that important ultra intelligence be disseminated to recipients without delay, and any other information was of insuffic ient importance to ju stify formal brie fing s. The typ ica l method by which uhra reached the recipients was as follows: Each moming the G-2 was briefed on the messages received during the night. If in hi s judgment the infomlation was important , the Commanding General and/o r the Chief of Staff were called in and the briefing repeated. Throughout the day, usually after dinne r and again after supper. the G-2 was briefed on later messages. Other recipi ents were briefed throughout the day whenever they found it convenient. At any time during the day or night an imponant message was passcd to the G-2 without delay. In the absence of the G-2, the uhra specialist was authorized to call any message directly to the attention of other recipients if the information therein warranted such action. The ultra spec iali st was ex pected to be cu rrent ly familiar with informatio n from other sources. but at no time was it desired that hi s briefings include a me rging of ultra and other information. Thi s distinction was closely drawn by the G-2, both for reasons of securit y and because the me rger of all sources of information was accomplished in other ways. At the Seventh Army the chief of the order of battle sec tion was a recip ient and the "opening up" of ultra was hi s primary responsibility. Estimates of the enemy situation were prcpared by still another officer in the combat intelli gence section , and ultra was injected into these estim ates of enemy capabiliti es by informal consu lt ati o n with the G-2. Target informat ion derived from ultra provided no problem for it had been agreed between the G-2 and the A-2 of XII Tactical Air Command, support ing the Seventh Anny, that all ultra targets wo uld be handled by the Air Force. With the order of ball Ie, estimates, and targets provided for in the above manner, the rcmai ning uhra req uired lillie o r no corre lation with other information. Attached hereto as Tab A is a statement of two instances in which ultra provided intelligence of extreme tactical imponance to the Seventh Anny. All rec ipients

152

SIGNALS INTELLIGENCE

we re genuine ly inte rested in ultra and aware of the treme ndous advantage provided by source. On one occasion the G-2 re marked, "You know, this just isn ' t cricket." Generally, ultra information is of primary value in a static or de fensive situation; th is is indeed true of all information and inte lligence derived the re from . In the attack and pursuit, inte lli gence has done its job durin g the planning phase of the operation, and subsequent information is usuall y outdated by the time it reaches the command . The examples in Tab A, however, cover both situations. 3. Security : It is difficult to really know the security situation at a command, but it was evident that some recipients at Seventh Ann y were e ither not cogni za nt of or suffic ientl y impressed with the need for securit y in the use of ultra. There was little questioning the minimum security requireme nt that the nature of source must not be disclosed, but thi s understanding did not ex tend to an appreciation that informat ion derived from source must be properly safeguarded and serve only as the basis for tactical orde rs when the informati on itself cannot be "opened up." The greatest misunderstanding here was in the idea of " cover." Reci pie nts were inclined to believe that cover was an invention; the idea of cover as an indication from an ac tual but open source was alien to a ll but a few rec ipients. The securit y situation at Seventh Army was complicated by the fact the G-2 of VI Corps, and at the outset the Commanding General , we re former recipients at a time when the Corps was operating at Anzio as a virtually independent task fo rce. On a few known and probably on many un know n occasions the G-2 o f Army passed on " hunches" to the G-2 of Corps; the latter was certainly aware of the ir orig in but was not bound by security regulation s. In thi s s ituation it is almost imposs ible to ex pect that a Corps G-2 , formerl y in the picture, will not receive ultra, thinly di sgui sed; as the Army G-2 frequently stated, "You can deny membe rship in the Elks to a person, but once he has been an Elk and forfeits his membership he will still reme mber the initiation." Generall y speaking, however, the handling of ultra from an intelligence point of view by the G-2 was consiste nt wit h security regulations. The difficulties at the command were more directly traceable to the Commanding General and in the operational use of ultra. The reasons for this are manifest. Charged w ith responsibility for success or fa ilure in the battle, the temptati on to knowingly or unknowingly emp loy ultra improperly is well -nigh irresist ible at times. This, considered a long wit h the fact that the Command ing Genera l is relatively isolated from daily security reminders by the ultra specialist appears to be the major cause of violations . . . .Once the violat ion was observed, the facts in the case were obtained from the G-2 and reported to London. The G-2 was never told by the ultra specialist, a lthough he probably guessed, that violati ons were being reported. 4. Conclusions: The policy of a separate channel and the assig ning of an ultra spec iali st to American comma nd s is basica ll y sound. If nothing e lse, the mere presence of a War Department representative has a salutary effect. For reasons of administration the SCU/SLU link might we ll become the responsibi lity of the s pe-

153

DISSEMINATING THE PRODUCT

cialist. There is often a real question whether a particular security violation is the responsibility of the SLU or the specialist. Moreover the command is frequently confused in their administrative relations with the SLU, and are never certain whether they should deal with the SLU or with the specialist. There was never any difficulty of this kind at Seventh Anny, but there is a real danger in the failure to provide unified control. There is much to recommend that all ultra mailers be considered within the province of the Commanding General and his Chief of Staff, rather than as an instrumentality of the G-2 . In the first place, the G-2 might then be subjected to controls which cannot be exercised under the present arrangement. Moreover, by attaching the ultra functions to the Chief of Staff's office, the ultra specialist has direct access to the Chief of Staff and the Commanding General and thus can more carefully safeguard security at the point where it is mos t vulnerable . Al so, by removing the ultra function from the G-2, security is improved by the mere fact that it is not automatically associated with intelligence. The reading and understanding of ultra security regulations is not sufficient to insure the ir inviolability. In addition to constant reminders by the ultra s pecialist. periodic warnings from high commanders seem desirable. The subject must be kept a live one at all times, and the interest of the War Department continuously impressed upon recipients . Regardless of what other persons are recipients at a command. it seems particularly desirable that wherever possib le the order of bailie chief be included. At Seventh Army this occurred through accident: the persona lity involved was "in the picture" at a higher headquarters before joining Seventh Army and his status was continued. But this might well become standard practice. other considerations permilling. for the order of bailie man is the key to opening up ultra and the preparation of sound strength estimates. Finally, it is essential that well-grounded intellige nce officers be assigned to commands. The proper interpretation of ultra is very often dependent upon a thorough familiarity with infonnation derived from other so urces. TAB A (I) Late on D plus 2, ultra infonnation reached the Seventh Army which indi cated that the enemy was withdrawing from south and so uthwe st France. The Commander was faced with a major decision inVOlving two factors: a. To what extent could the enemy be pursued and outOanked. considering the difficulty of supp lying advance columns over a beach, and b. Would the enemy counter allack on the right flank , from the Maritime Alps, and so endanger rear communications.

Ultra provided the answer to the second question, at least a suffi cient answer to pennit a calculated risk: there was no indication that the enemy would adopt an

SIGNALS INTELLIGENCE

154

attitude other than defensive on the Oank. Accordingly, it was decided to pursue and all unloading priorities were altered, with the whole emphasis given to fuel and vehic les . Also, Task Force Butler, which had penetrated deep in the enemy rear, was reinforced with the 36th Div. Together they establi shed a road block at Montelimar in the Rhone Valley and cut the enemy escape route. The road block proved insufficient to destroy the 19th Anny and they succeeded in fighting their way out, but all heavy equipment was lost in the process. It is also noteworthy that ultra guided public relations all through these opening days of the campaign, for it was clear from source that the enemy was not aware of the character of the forces operating in hi s rear. He apparently believed that there were only guerilla forces endangering his lines of communications; and so the existence and operations of Task Force Butler were not disclosed to the press. (2) ... Shortly after the German Eifel offensive in mid-December, it became evident that that the enemy was preparing an offensive in the Saar-Palatinate, possibly in conjunction with a coordinated attack out of the Co lmar bridgehead and a crossing of the Rhine N. of Strassburg. Gennan Air Force reconnaissance orders from ultra pin-pointed the possible crossing area but source was relatively quiet on the Co lmar attack; open sources, however, clearly suggested the latter intention. But it was in the sector between Saarbrucken and the Rhine that ultra provided information necessary for a proper estimate of the situation and preparations to meet the attack when it came. During the critical days prior to the attack the Flivo with the German First Anny was being read consistently, and enemy order of battle and boundaries were throughly known. It was possible, therefore, to state with relative certainty that the main effort in the attack wou ld be made west of the Hardt Mtns., wi th a secondary attack between the mountains and the Rhine. If there was ever an essential e lement of information this was it, for the passes through the Vosges Mtns. were a serious obstacle to the rapid movement of Seventh Army reserves. Based on this appreciation, the sector west of the Hardt Mtns. was reinforced with the 2nd French Ann'd Div. and the 36th U.S. Inf. Div., both moving from the Rhine plain. It shou ld be noted that up to the day of the attack , German intelligence failed to pick up these moves. When the attack was launched on I Jan .. the German main effort collapsed completely. Their on ly success was in the sector of the secondary effort, in and East of the Hardt Mtns. This German offensive was properly appreciated and preparations made to successfully meet the threat. Lacking ultra it seems very doubtful whether the attack would have been repulsed , or whether other sources of information would have given advance warning. Open sources provided only the

most meager evidence of an attack, and there was much opposing evidence suggesting preci sely the opposite - a thinning out in the sector and movement of units away from the Saar-Palatinate to reinforce the North.

***

DISSEMINATING THE PRODUCT

155

From SRH 032. This reparl was s /lbmil/ed by Major J ohn H . Gllnn, Ih e Special Se c llrily Officer assi!?ned 10 General Do ug la s MacArlhllr 's SOlllhweslern Pacific Area headquarlers. As GLIIln illdicales, MacAl'lhur ran his Ih ea ler like a pri lla le fiefdom, and lhe members of his slaf[ were nOl always happy 10 deal wilh people Ihey regarded as illleriopers from Ihe War Deparlmenl.

WAR DEPARTMENT MILITARY INTELLIGENCE SERVICE Washington 25, D.C. 23 October 1945

MEMORANDUM FOR THE SPECIAL SECURITY OFFICER, MIS: SUBJECT: Report of Major John H. Gunn, 0-388247 The unde rsig ned office r departed Wa shin gton on 15 Se pt embe r 1944 for Brisbane, AUStralia. I reached Bri sbane o n the night of 22 September. At thi s time most of GHQ was in Ho ll andia, Dutch New Guinea, in the final stages of pl an ning and preparation for the invasion of Leyte. The Leyte operation had been accele rat ed and it was necessary that those SSO 's designated to serve the participating commands cut short their indoctrination course in Bri sbane and proceed to Hollandia at an earlier date than planned. My principal duties in Holland ia were to assist Ll. Colonel Ashby, then Special Security Representati ve for the theater, at his headq uarters. I had been introduced to Ll. General Sutherland, Chief of Staff, and it was indicated that I was to accompany the Advanced Eche lon into Leyte and serve Genera l MacA rthur, Ge ne ral Sutherland, and appropriate G-2 and G-3 personnel. The acceleration of the date for the Leyte operat ion necessi tated Ashby's being absent several days while placi ng SSO's at the subo rdinate commands. During thi s pe ri od the undersigned was left at G HQ. I was impressed wit h the fact that the people on the working level in the theater are, despite the idiocyncrac ies, j ealousies and animosities of their superiors, an amazingly cooperati ve group. The Ch ief of Staff had indicated that there wo uld be no room on the Command Sh ip for an SSO. As a result he referred Ashby and me to the G-3 secti on to make arrangemen t for tran sportati on to Leyte in some oth er vesse l in th e Advance

Echelon convoy. The G-2 offi cer in c harge of the detai ls of loading the G-2 section was to notify me of the proper time and means of dcparture. I am confident that this charge never agai n crossed this offi cer 's mind, for Ashby and I discovered the follow ing Saturday night that the G-2 convoy had loaded without any notice to o ur

156

SIGNALS INTELLIGENCE

office. In the improvision that was necessary I finally arranged to " hitch hike" to Leyte aboard a 6th Anny convoy leaving the next day. The trip to Leyte was uneventful. I carried no classified material other than a black handbag filled with cryptographic material. This material I kept stored in a locked closet in the ship's captain's cabin. This seemed the most secure procedure aboard a vessel whose Commanding Officer kept 18 quarts of bourbon whiskey in his ship's safe, and kept his top secret battle plans for several operations on top of the desk in hi s sleeping quarters. We arrived off Dulag at the southern portion of the invaded beach area three days after the initial landings. The usual confusion prevailed on the beach and I was at a loss to find a secure place to keep my cryptographic materials until I could find transportation to GHQ, then located at Tacloban, 20 miles north. Such transportation was not immediately available. After two rather unpleasant nights ashore, when minor but apparent enemy infiltrations of the beach position occurred, the undersigned decided that the most secure place for himse lf and the classified material was aboard a boat in the harbor. A problem like secure storage seems minor, but it is typical of the practical problems our system often poses. Upon arriva l at GHQ I presented myself to the G-2 section and commenced operations. I was introduced to General MacArthur on the first day of my arriva l and found him at all times as cordial and cooperative as possible. It is a matter of record that the same observation cannot be made of a number of his immediate subordinates. It took only a short while to comprehend the low state of security at the headquarters. For example, no adequate safe acco mmodations ex isted, and the loose physical security of the headquarters was evidenced by the presence of innumerable Filipinos who stalked in and out of the staff section area at random. As for the handling of Ultra, complete insecurity was a recurring ma lignancy. The details of numerous security violations involving our material that occurred during this period have been reported to MlS. One fact was obvious-the main factor contributing to such incidents was a basic lack of appreciation of the val ue of our material. The principle of sacrificing tact ica l advantage to securi ty was not even accorded lip serv ice. It seems incred ible that Ultra could be obtained and passed by a non-recipient to a non-authorized recipient over a radio relay telephone wh ich was almost certa inl y being tapped by the Jap. This, and other occurrences as serious. if not as spectacular, are indicative of the security point of view of the persons des ignated to

receive

OUf

material.

Without going furth er in detail, there are certain conclusions that I ha ve drawn with regard to the si tuation as it occurred on Leyte. First, General MacArthur has been overso ld by hi s subordinates on the quality of the production of his theater agency, Central Bureau. For example, on one occasion General MacArthur expressed to me a desire to see more diplomatic material , as " I get all the other infomlation anyway."

DISSEMINATING THE PROD UCT

157

Second, tacti call y, the Navy mate rial was incomparabl y supe ri or to Army production, For example, no Arm y material was o perationall y comparable to the Navy material describing the pending Japanese reinforcements in the Ormoc area, Third, there was no excuse for relaxing e nforcement of o ur regulations to permit passage of Ultra down to Corps level. In no case can I recall seeing any Ultra appropriate for a Corps in its limited sphe re of ac ti on, and I can imagine no case where a Corps, operating non independentl y, cannot be served by an operational order, Fourth, none of us, staff officers or SSO's, dealing with Ultra, were adeq uately indoctrinated in the handling of thi s material in the field , and in the security practices necessary to ensure its safety, The G-2 me n lacked any adequate securit y indoctrination, In the future the indoctrination of appropriate personnel should be as integral a part of the ir intelligence trai ning as the preparation of maps or summaries, The security qualification should be a stric t exaction for the holding of a responsible intelligence pos ition, It was not in G HQ, SWPA, Nor were the SSO's adequately prepared for the ir job, Extensive training in the drafting of operational orders, as a security dev ice, is one fundamental necessity in which no early SSO rece ived any training whatsoever, The unde rsigned, without any previous experie nce, drafted several suc h orders in efforts to present an acceptable substitute for an otherwise unwarranted dissemination , It would have been an infinitely more simple task had the unde rsigned even seen an o pe rational order before he tried his hand at drafting one. Further, if the SSO is to work in an Intelligence Section , he should be thoroughl y schooled in the G-2 duties he will perform, Camp Ritc hie should furni sh the best answer to this problem . Fifth, closer coordination between Army and Navy Ultra rec ipie nts and communications personnel is essential. Tales of the Navy "i ndi scre tions" in SWPA are leg ion, but in a great measure they may be attributed to a lack of mutual unde rstanding and agreement. For example, Navy furniture Shipped from Hollandia to Leyte was addressed to "the Ultra section. " and mess bills were addressed to Lt X, c/o the Ultra section. This came about in great measure because the Navy conside red "U ltra" as a classification , not a code word. Similarly, Nav y acceptance of SSO pe rsonnel as participating counterparts aboard ships receiving Ultra is essential to secure handling of the mate rial - si nce Navy Ultra may obviously be used as the basis for ArnlY directives to Army units. On such directi ves a trained SSO should be infinitely more authoritative, security w ise. than a Navy communica tions man. On completion of the transfer of GHQ to Leyte, I was sent to the Office of the Deputy Speci al Security Representative in Brisbane. The re is no need to include in thi s report any description o f the organization and work of this office, .. ,

*** JOHN H. GUNN Major, Inf.

CHAPTER VIII The Politics of COMINT The motto of the great Army cryptologist William F. Friedman was " Knowledge Is Power." The power derived from the ability to exploit high level s ignals intelligence meant that access to and control of COMINT became a political football in various bureaucratic and partisan struggles for advantage. Within the Army, the Military Intelligence Service fought with the Signal Corps over control of the Signal Security Agency. Excerpts from SRH 141-1 2nd SRH 141-2 document the course of the dispute, which was finally resolved (more or less) in December 1944, when the Military Intelligence Service gained operational control over Arlington Hall. On another bureaucratic front , the Military Intelligence Service tried to take on the Navy, claiming that Navy cooperation in sharing signals intelligence was lacking. Another selection from SRH 141-2 reveals the extent of the problem and the fierceness of the rivalry. Finally, in SRH 043, Comments on the Dewey-Marshall Exchange, Colonel Carter Clarke, Chief, Special Branch , describes what happened when Army Chief of Staff General George Marshall tried to ensure that COMINT secrets would not be compromi sed in the course of the 1944 presidential election contest between incumbent Pre s ident Franklin D. Roosevelt and Republican challenger Governor Thomas E. Dewey of New York.

160

SIGNALS INTELLIGENCE

From SRH 141-1. Two months after joining Stimson's staff to recommend

better measures for handling signals intelligence, Alfred McCormack had become convinced that the Military Ill/elligence Division should be given overall cOll/rol of all ill/ercept activity. To help make his point about the importance of secret intelligence, he makes reference to two intercepted Japanese messages dealing with this subject, one from Japanese Foreign Minister Yosuke Matsuoka. and one from from his successor, Foreign Minister Shigenori Togo. March I, 1942 MEMORANDUM FOR: MR. McCLOY Subject: Signal Intelligence Service Here are noted down the points discussed with you, touching on the question whether it is sound organization to leave with the Signal Corps the job of collecting and selecting intell igence materiel of great importance, instead of lodging that responsibility with G-2 as a part of its over-all job of collecting and supplying intelligence: I. It appears to me to denote a rather negative conception of intelligence to have G-2 simply taking what the Signal Corps catches and turns in, leaving with the Signal Corps the responsibility for deciding how much it will catch and tum in, and from what sources. I am sure that the Signal Corps listens seriously to all suggestions from G-2, but the power of decision remains in the Signal Corps, and there is nobody in G-2 who has an affirmative responsibility for seeing to it from day to day that the material collected and processed (a) is all that possibly can be collected and processed and (b) is derived from the sources that are at the time of greatest importance from G-2's ang le. Intelligence activities should be directed affirmat ively and vigorously. Compare Minister Matsuoka 's direction that "our principal aim is to collect ellery possible bit of intelligence concerning the United States" (18846) and Minister Togo's reference to the "absolute necessity of securing intelligence of this sort" (i.e., the sort not obtainable from newspapers and the radio; 27500). 2. The men in G-2 who are working on the processing of intelligence are too remote from the collection of this material. Let us suppose that Col. Heard, [text withheld] decided [text withheld I or was instructed by his superiors, to gather in as much current material as possible on political conditions and diplomatic maneuvering in [text withheld] .... He would not get any such material from the source under suggestion, though the S.1.S. has a lot of [text withheld] material on its shelves, all in its original form, which has never been tackled because of lack of personnel. If there were one officer in G-2 whose responsibility was to keep up to date on G-2's needs and be ready at any moment to tum the S.I.S. to work wherever the heat was

THE POLITICS OF COMINT

161

on, and if that officer could give orders to the S.I.S., maybe in such a situation all hands not elsewhere essential would go to work on the [text withheld] material. At least, there would be an evaluation of relative needs, and a conclusion would be arrived at in the light of all the facts . Moreover, the real needs as to personnel , equipment and supervision would quickly become apparent if that sort of an evaluation had to be made once in a while. 3. This is a situation in which an infinite number of little things are important. From the dim perceptions that I have obtained to date as to how an Army works (my past military observations having been from the bottom up), [ have a notion that, if a good idea about a matter of detail should originate in one of the sections of G-2, affecting what should be done somewhere down in the ranks of the S.I.S., by the time it got up the G-2 command, over to the Chief Signal Officer and down through the appropriate channel , the impulse behind it might have become so attenuated as to be insufficient to actuate a full-grown adult. This is not meant critically, since, as I have already told you , I have been impressed by the ability of all the officers that I have encountered to date. 4. Under the present system of divided responsibility there is nobody in G-2 who knows, or could know, what goes on at the various S.LS. stations or what decisions made from day to day at those stations affect the content of intelligence coming from those sources . Do we know, for instance, that we might not have obtained a lot more Honolulu material except for some decision by the Signal Corps in routine matters, e.g., to shift an experienced operator elsewhere, to postpone repairing a broken antenna, to use particular receiving apparatus for some other purpose, to run a station less than 24 hours a day, to shut it down for repairs when X might be heard, rather than when Y might be heard? These are minor items, but they might have an important effect on the content of intelligence. Yet in the nature of things the men responsible for intelligence could never know about them. 5. As to items of major importance, it necessarily must be true that the S.I.S. is in competition with the manifold other activities of the Signal Corps for money, supervisory direction and personnel. It seems illogical to keep one intelligence unit separate from all other intelligence units, and in competition with units not concemed with intelligence. The production of intelligence from this particular source ought not to compete for money, direction or personnel with anything except the production of intelligence from other sources. Then you could shift your money and personnel into the branch of intelligence which seemed most likely to be fruitful. As it is, it might be that the Signal Corps would be spending much too much time and money on the production of intell igence material from thi s source, in relation to the time and money spent on the production of intelligence from other sources. On the other hand, there is some ground for thinking that the Signal Corps is not spending nearly enough time and money on the production of the type of intelligence here under discussion.

162

SIGNALS INTELLIGENCE

6. Under the conditions which exist, it is necessary for the Signa l Corps to make many choices which affect the content of the intelligence gathered. It must dec ide at each station what to listen to; and doubtless there wo uld sometimes be a wide choice. Then it must frequently dec ide the importance of a particular item, in determining what ri sks may be taken in getting it to Washington promptly. The traffic that comes in is much greater than can be handled ; and at each ste p in the process of putting the raw material into fini shed form there must be a selecti on of a small number of items out of a relative ly large number of ite ms. The average number of items going into the mill each day is understood to be around 1,200. Of these, only 200 are read (by the junior officers in charge of the various sections) and only 60 go beyond the stage of readin g. In a particular situation-such as the French si tuation as I have described it to you-we get out hardl y more than a sample of what goes in , and a mere fraction of a sa mpl e of what mi g ht go in. Consequently, the choices here are very important, especially since the material that comes out must be examined with two questions in mind , ( I) whether it is valuable in itself and (2) whether it suggests that other material from the same source should be obtained for examination in larger quantities. 7. The present division of functions appears to increase quite materially the number of persons familiar with this material or with the fact that it exists. Without having heard such arguments as there may be to the contrary, it appears to me that the organization would be sounder, and the job more effectively done as an intelligence job, if the S.LS. were made a part of the General Staff, G-2 . A. McC.

THE POLITICS OF CO MINT

163

From SRH 141-2. The issue over whether the Signal Corps or the MilitOlj

Intelligence Division should control SSA came a large step nearer resolution when Assistant Secretary of War John 1. McCloy accepted McCormack's arguments in favor ofG-2 control. This memorandum was sent to Lieutenant General James 1. McNarney, the Army's Deputy Chief of Staff. As deputy to Marshall , McNarney played a major role in Army reorganization. (To oversimplify, Marshall fought the war while McNarney ran the Army.) The Generals Ingles and SomeI've II mentioned in the memorandum were, respectively, Major General Harry C. Ingles , Chief Signal Officer; and Lieutenant General BrehDl! B. SomeI've II , Commanding General, Army Service Forces (ASF). During World War/l , the Army Service Forces commanded all of the Army's technical services, including the Signal Corps.

22 August 1944 MEMORANDUM FOR GENERAL McNARNEY: I have thought a great deal about the matter of transferring the Signal Security Agency to the Military Intelligence Service under the Assistant Chief of Staff, G-2 . I have given it renewed consideration after the receipt of General [ngles's memorandum and General Somervell's indorsement. I. It was my impression that the merits of thi s transfer were considered by the committee which was set up to go into the reorganization of G-2 and that we came to the conclusion that if the Assistant Chief of Staff felt that he needed this agency to render the proper service it should be given to him . Subsequently he delivered a memorandum to you which as I understand it was a proposed outline of the procedure to accomplish the transfer rather than any argument in favor of it. The memoranda of General Somervell and General Ingles do not criticize the procedure outlined but object to any transfer whatever. 2, The chief stated objections in Ingles's memorandum seem to be as follows: a. That the MIS is only a "staFr' section and not an operating agency and that it is not "authorized" by Congress. b. That for some reason not clearly stated the Signal Corps will not be able to furnish the MIS the same communications facilities and services that it provides for the SSA. c. That personnel engaged in operating communications facilities will not be able to make proper use of codes and ciphers if such codes and ciphers are supplied by the MIS rather than by the SSA. 3. As to the first of these arguments, above, I do not agree that the MIS merely because it is under the supervision of the A.C. of S., G-2 is not an operating agency. [n fact, [ think it is, and should be. Its relegation to a minor role between World War I and II was a tragic mistake and certainly the Signal Corps was not

164

SIGNALS INTELLIGENCE

able to keep up any vigorous or broad system of communication intercept during this period . 4. In my judgement one of the chief pillars of our national security system after the war must be an extensive intercept service. If we are to be a military power or, indeed. if we are to play an active role in world affairs, we cannot afford to leave thi s field entirely to the British and the Continental powers. It is one of the best sources of intelligence that there is and I would take it out of any existing service agency immediately in the hope and belief that it would develop into an organization which would stand a better chance of perpetuation in peacetime. 5. I think that the establishment of such a vigorous agency would be an attraction to the Intelligence Service. The curse of our so-called Intelligence Service to date is the attachment to it of only those officers who have social acceptance and means enough to enable them to pursue a life of relative ease. If we try to put this thing now on its proper level and scope it may be possible to create a real field of activity for ambitious and capable officers. I do not think that we will ever get anywhere if we merely seek to maintain a system whereby direction comes from the MIS and operation lies in the Signal Corps. It ought all to be under one roof. The British have been more successful in this field than any other country, I think. Certainly our effort s to date have been most fragmentary as compared to thei rs. They use the system that MIS would wish to follow and 1 think it is the most persuasive example we have to look to. 6. 1 do not follow the argument on codes and ciphers. Certainly they can be separated from the operation or communication facilities . In fact, such a separation now exists in the Signal Corps and I do not see how it would be more difficult for the MIS to turn over codes and ciphers to the users than it is for the SSA to do so now. 7. I do not envisage that this transfer involves MIS responsibility for all Signal Corps training. Signal Corps will continue to provide signal training as before the proposed tranfer and the MIS will obtain personnel trained by the Signal Corps to the same extent that SSA now obtains such personnel and in the same manner that other components of the Army obtain such personnel. There will be certain highly specialized training in the Signal Intelligence which will be provided by the MIS but I think this is all to the good as it will give added emphasis to the intelligence aspects, which have been neglected heretofore. 8. I do not believe that it is proposed to " tear to pieces" any existing organization but to transfer SSA to the MIS in its present form and thereby to effect a closer integration of its activities with those of the present MIS. 9. I do not believe that it is good practice to have the MIS by general directive and the establishment of general policies to attempt to create a workable intelligence organization in the low echelons of the Army Service Forces as they are now attempting to do. If it has not already occurred, I am sure that there will be occasions when the MIS will seek to direct the SSA into certain activities or lines as to which, if the SSA is inclined to disagree, they may appeal to the higher levels of

THE POLITICS OF COMINT

165

the ASF. When this occurs, there is no one short of the Chief of Staff who can adjudicate the point. 10. Finally, I think that the argument that the present arrangement is successful because SSA furnishes the bulk of the information that the MIS has available is unsound because the total Intelligence available to the MIS really covers a much wider field and actually only a relatively small part of that intelligence is made available by SSA. Much of the most valuable work that the MIS have applied to the intercepts has come in the processing which that service has done on the material furnished it. No part of the Signal Intelligence which MIS has available in relation to European military matters is furnished by the SSA; no part of Naval Signal Intelligence is furnished by the SSA; and no part of the clandestine European field is supplied by SSA. About 70% of the diplomatic is and about 80% of Japanese military is. It is true that it provides the greatest amount of information obtained from anyone source. II. In summary, I believe that the MIS should be given the opportunity of proving that it can exploit this entire field to greater advantage. The enthusiasts are in the MIS; there is some great talent there, and I believe it is a source upon which a broad and effective Intelligence communication service can be erected after the war. The Chief of the Army Communications Services and the Chief Signal Officer are primarily interested in operating a signal communication service; they are not primarily concerned with Intelligence . I believe that the possibilities of greater accomplishment lie in the system advocated by the MIS than in a continuation of the existing system. I am not sure that MIS will take full advantage of all the possibilities but on balance I do believe that they would have a rar better chance if they had full control. By such control there is a greater likelihood that we will come out of this war with an organization that can be maintained through the lean years of peace. I would be quite certain that if Signal budgets were cut and SSA were in the Signal Corps, the latter 's budget would go, if not first, at least before the regular communications appropriations were cut. We will need an organization to support it whose primary and sole objective is Intelligence. Of course, there is no merit to the argument that there is no statutory authority for the MIS. First, I think there is sufficient authority to meet any need and, second, when can the ASF complain about no statutory authority? /s/ JOHN J. McCLOY J.J. McC.

***

166

SIGNALS INTELLIGENCE

From SRH 141-2. In this memorandum, the Special Branch takes on the U.S. Navy.

WAR DEPARTMENT MILITARY INTELLIGENCE SERVICE WASHINGTON 4 March 1944 MEMORANDUM FOR GENERAL BISSELL: Subject: Anny-Navy Agreement regarding Ultra. A. Background oj Army-Navy Relarions in UlrraJield. I. Except for the Japanese Anny material that recently has started to flow in, the Navy has had the full benefit of the work done by Arlington Hall and the Special Branch. It receives copies of the decoded diplomatic and related traffic. Through the " Magic" Summary it is furnished with all evaluations and studies made by Special Branch. When the British European sources were opened up, and the Military and Naval Supplement was established to report intelligence from those sources, arrangements were made to furnish this daily report to Cominch. 2. Little or no cooperation was received from the Navy in return. It did furnish certain infornlation about Japanese merchant shipping, and transmitted some Order of Battle infornlation to G-2 through other channels. In general, however, the principle was applied that cooperation in the Ultra field would be restricted to the technical and cryptanalytic level and would not extend to intelligence. Not once, during the entire period over which the Navy has been receiving the " Magic" Summary, has any agency of the Navy volunteered any infonnation that might have been of aid on any of the subjects of study covered by the Summary, although on some of them (such as Axis blockade-running) the Navy has had considerable infonnation at hand. 3. When the Japanese Anny source was opened up, it was decided not to furni sh the traffic to the Navy without some arrangement for reciprocity. For almost a year Arlington Hall had worked on Japanese Anny ciphers, had gradually solved them and applied itself to the tedious job of reconstructing code books, while Op 20-G, exploiting a parallel source and being in possess ion of intelligence that would have been of great value to Arlington Hall, withheld this intelligence and restricted its liaison to the technical level. In the opinion of the writer, the noncooperation of the Navy retarded the exploitation of Japanese Anny sources by not less than 6-months and resulted in depriving both services of intelligence that would have been of important operational use. 4. In spite of the Navy 's attitude, the writer dId not feel justified in wit holding

Brig. Gen. Caner W. Clarke presents Col. Alfred W.McCormack wi th the Di stingui shed Se rvice Meda l al the end of Wo rld War II. The two men were the arc hitecb of the Special Branch. Military Intell igence Service. whi ch eval uated and disseminated AnllY commun ications intelligence. (NA RA )

G e rman s ignal troop s operate an Enigma cipher device in the field durin g World War II. Th e Enigma was used to enc iphe r the commun ications of all three German armed se rvice s. (NARA)

Headquarters of the British Government Code and Cypher School at Bletchl ey Park . Here. British cryptanalysts broke the secrets of the Enigma. (NSA )

Co l. Telrord Tayor was appointed to head the London o rfice or the Special Branch in 1943. A prominent lawye r in c ivilian I ire. Tay lor went on to become a military prosecu· lor. Here. Brig
America 's lOp military commanders were able to base their plans on high-level intelligence derived from ex.p loitation of enemy com muni cations. Above. General Dwight D. Ei senhower addresses paratroopers. (NARA)

Ge neral Dou gla s MacA rthur. le ft , used s ignal s inte lli ge nce 10 bypa ss s uperior Ja pan ese forces in hi s is land - hoppin g campaign. (NARA)

To prosecllte its war against the U·boa t. the U.S. Navy erfecti vely used cryptologic intelli· ge nce. but it was initially relu ctant to s hare thi s source wi th the Anny. (NARA)

AmlY Chief of Staff General George C. Mars hall mee ts wit h New Yo rk Governor Thomas E. Dewey in 1943. To prevent any compromise of crypt
World War II tac tica l signa ls int clli gence units. Left: radio direction-finding personnel of the 21 s t Signal Service Company's Radio Inte rcept Platoon in Newfoundland. Be lo w: analysts of Signa l Security Detachmcnt "D" in England, (N ARA, INSeOM)

Intercept personnel of the 128th Signal Radio Inte lli gence Company work out of a commandeered villa in the Mediterranean Theater of Operations. (lNSCOM)

A rad io direction-finding site of the 126 th Signal Radio Int elligence Company at Brisbane. Aust rali a. (NARA)

Army Chief of Staff General of the AmlY Dwight D. Ei se nhower at Arlington Hall Station shortly after World War II . Other persons in the picture are, from left to ri ght. Maj. Gen. Hoyt S. Vandenberg, Army Assi stant Chief of Staff G- 2; Brig. Gen. PresIOn W. Corde nnan , Chief. Army Security Agency: Col. Frank B. Rowlett ; and William F. Friedman. (NSA)

THE POLITICS OF COMINT

167

from the Navy infomlation, derived from Japanese Army traffic, which might be of value to Naval operational units. Therefore,with the approval of the A. C, of S., G2, an arrangement was made for forwarding promptly to the Navy all information on which operational action might be taken. 5. The Japanese Army traffic is exploited by British units in England and India and by an Allied unit in Australia, as well as by Arlington Hall. As the work of Special Branch on this material progressed, it became apparent that we were in a less favorable position to get intelligence out of it than were the British, because they had the benefit of the U.S. Navy traffic and intelligence, whereas we did not. Efforts were made, by personal conversations with representatives of the Navy, to bring an end to this disgraceful situation. Those efforts probably contributed to bringing about the Navy 's proposal for an exchange on the intelligence level. However, the success of Arlington Hall in opening up the Japanese Army sources was what probably brought the matter to a head. 6. In the discussions involved in preparation of a written agreement, the Army was represented by the writer, Colonel McCormack and Major Perdue of the Special Branch. Following the first meeting, the Special Branch submitted a simple agreement for an exchange of traffic and intelligence, along the lines of the agreement which exists between the A. C. of S, G-2, and the British GC&CS. This was not acceptable to the Navy, who insisted that liaison officers be exchanged and that various restrictions be put into the agreement, pursuant to which, in effect, either party might give or withold intelligence as it might see fit. 7. The final draft of the agreement, as prepared by the Navy, was not acceptable to the officers representing the Army, without a statement by the Navy of the principles that would be applied in interpreting the agreement. Those principles were stated on several occasions by Captain Stone, and were generally to the effect that the restrictive provisions of the agreement were necessary to cover special cases, which might never arise; and that the basic principle would be cooperation to the fullest extent in furnishing the Army with all the "Ultra" material that it might find of value. It was on the basis of such assurances that the agreement in its final form was submitted to the A. C. of S. , G-2, with the recommendation that it be approved by the Chief of Staff. 8. When the Special Branch appointed its liaison officer and he started to go through the Navy daily matarial , it was at once apparent that no one officer could sit in Op 20-G and make a final determination of what traffic might be of intelli gence value to the Army; and that the arrangement would work effectively only If the liaison officer were pennitted to select decoded traffic and forward it to the Special Branch without being required to demonstrate how and why it might be of value. The decodes are readily available on duplicate cards and can be handled conveniently by the various specialist groups within the Special Branch who may be working on the subjects to which they relate or might relate. 9. The experience of the Navy's liai son officer with the Special Branch was

168

SIGNALS INTELLIGENCE

somewhat similar. He discussed the point with the Deputy Chief of the Branch, and was informed that an ex tra copy of all messages wo uld be obtained for hi s exclusive use and that no restrictions would be put upon him in respect to what he might select; that if he cared to do so he might take all the traffic. 10. A day or two after the Army liaison officer, Major Snow, had started to function , he was informed that there would be certa in general restrict ions upon what he might transmit to the Special Branch. Those restrictions we re then embodied in a memorandum, a copy of which is annexed as TAB A. B . Proposed Resrricrions. (i) U.S. Novy operarional disparches and reports. II. It is not clear just what this covers, but it appears to mean that we cannot receive the CINPCAC Bulletin, despite the fact that it is given to various British intelligence organizations, and despite the fact that one of the purposes of the agreement was to put an end to the situation in which the Briti sh were receiving material from the Navy which the Army did not receive. (ii ) Japan ese j1eetmovemel11s and locations. 12. Major Snow was instructed to select and transmit to Special Branch only those items relating to the Japanese fleet as represented major concentrations or movements. Insofar as the fleet itself is concerned, the daily Summary received from Captain Smith-Hutton will probably be sufficient to supply all major concentrations and changes. However, the Navy classifies hospi tal ships, tankers and auxiliaries as part of the fleet. Our information covers these vessels also; and to deny us the correlating information is to impair the value of our own intelligence materi al. 13. Moreover, communications dealing with the fleet, or elements of the fl eet, ma y be of interest and value in a g reat man y incidental ways; and no one can deternline what value they may have until he has received and studied the material over a period of time. (iii ) Japanese N{H'al Air irems. 14. No one dealing with Japanese Air intelligence can work simply on Army or simp I yon Navy material. An accurate picture of Japanese Air strength , dispositi ons, production, wastage, etc., can be obtai ned on ly by keeping up with both services. We do not intend to give the Navy Air material detailed processing, but we do want to have it at hand for the purpose of familiarizing ourselves wit h the type of material and spotting major pieces of information. (iv) OPTELS. 15. The significance of this limitation is not understood. It is ass umed that the OPTELS are received from the British Admiralty. If they bear any relationship to the Ult ra problem in the Pacific they should be included in Ihe exchange. C. Proposed scope of marerial ro he released by Navy. 16. Op 20-G's memorandum (TAB A) provides Ihal Ihe malerial 10 be released to the Army "shall ih- c1ude on ly thai which concerns the Japanese Army." Apart from the well-proved principle that intelligence of value can often

THE POLITICS OF COMINT

169

be obtained from messages bearing no apparent relation to the subject of interest, the restriction is believed to be unsound, because the war against Japan is not being conducted by Navy against Navy and Army against AmlY. Where the U.S. Navy operates it is fighting the Japanese Navy and Army, and the U.S. Army is fighting the Japanese Army and Navy. The Army traffic and the Navy traffic are interrelated parts of a sum total of Japanese military communications. They cannot be separated without losing an important part of their potential intelligence yield. D. Proper /Ilferprefafioll of Agreement. 17. There is only one sound basis for exchange of cryptanalytical intelligence, and that is to permit the consumer to determine what may be of value to him. This is the system under which the British operate, and it accounts in large measure for their success in this field. In the Watch Room at Bletchley Park, representatives of each of the 3 Services determine for themselves what may be of interest to the Ministry and field Commanders whom they represent. So much emphasis do the British put upon this point that , after all the operations of Bletchley Park have been completed, every piece of paper produced, in the way of transmitted messages, paraphrases or reports, is put through a so-called "Intelligence Exchange" where a representative of each Service checks the whole output, in order to make sure that his Service has not been omitted from the distribution list for any paper that it might find of interest. 18. There may be special cases in which a particular message may be of such a dangerous character, from the angle of security, that the originating Service might not want to give it any circulation at all. Such cases are provided for by the last paragraph of the Army-Navy Agreement. Such cases, however, should be very rare. 19. To be of value to the Special Branch, the liaison arrangement with the Navy must permit the liaison officer a wide discretion in the selection of material for transmission to the Special Branch. If only for the purpose of educating the personnel of the Special Branch dealing with particular subjects, it is necessary at the outset that he select everything except items which are obviously too trivial or which have to do with very specialized subjects such as promotions of minor Naval Officers, from which the Special Branch would probably not obtain anything of intelligence value. This was the understanding under which the liaison arrangements were worked out and the arrangement recommended for approval. If the Navy is unwilling to commit Itself to this principle, then it is recommended that steps be taken to abrogate the agreement and return to the status quo. 20. Restoration of the status quo would not be a desirable solution of the problem because it is certain that the adding up of Army and Navy radio intelligence would produce a sum total of intelligence greater than could be produced by the 2 organizations working separately. Such a solution at this time, however, would not hurt the Army as much as the Navy 's noncooperation has hurt us in the past, because we are rapidly gelling to the point where we will have an independent basis for intelligence in the Pacific and the Far East, and at the present rate of

170

SIGNALS INTELLIGENCE

progress we could struggle along without the Navy 's cooperation. 21. If the Navy is unwilling to accept the interpretation of the Agreement which we consider reasonable, and which was our basis for recommending it, then it is believed that all cooperation with the Navy in the radio intelligence field should be suspended . Specifically, it is recommended: a. That the Signal Security Agency no longer furnish intercepted Navy traffic to the Navy; b. That the service perfonned by the Signal Security Agency for Op 20-G, in collecting and transmitting weather traffic, be discontinued on 30 days ' notice; c. That exchange of traffic in the diplomatic field be suspended; and d. That the Special Branch discontinue furnishing the Navy with its reports and studies. E. European Intelligence. 22. The Navy has never been willing to ag ree to any exchange of radio intelligence affecting the European Theater. It receives what is provided by the Special Branch, but it contributes nothing from its own sources or from the British Naval sources. Further than that, the Navy has refused to pennit the British to furnish the War Department with any intelligence from British Naval sources. 23.When the Navy first proposed to do business on the intelligence level, it was our impression that the agreement would be general, covering all theaters of operation. It developed, however, that the Navy was unwilling to agree to any exchange except for the material relating to Japan. It was suggested by the Navy negotiators, however, that once the first step had been taken and a suitable arrangement worked out for the Japanese traffic, an agreement covering other areas might also be worked out in due course. 24. In view of the attitude now taken by the Navy, it is apparent that they will do business on ly when they have something to gain, and that there will be little prospect of extending the existing agreement to the European Theater. It is therefore recommended, independently of whether a modus vivendi is arrived at for Japanese material, that Special Branch be instructed to discontinue furni shing to Com inch any intelligence relating to the European Theater. 25 . It is now apparent that the Navy proposes to do business at ann's length. We should accept that attitude and act accordingly, giving the Navy nothing which our agreements do not require us to give and holding strictly to the letter of the agreements. After a year and a half of dealing wi th the Navy in this field , it is my conviction that you cannot do business with them in any other way, and that you will get more in the long run if you take this viewpoint. CARTER W. CLARKE Colonel, General Staff Chief, Special Branch, MID

171

THE POLITICS OF COMINT

From SRH 043. In the fall of 1944, Army Chief of Staif General George C.

Marshall became concerned that the fact the United States had been reading enciphered Japanese traffic prior to the Pearl Harbor allack might be disclosed in the course of political debate in the forthcoming presidential election campaign. He wrote a personalleller to Republican presidential candidate Thomas E. Dewey urging him 1I0t to make any statements tizat might compromise American successes in cryptanalysis. Portiolls of this leller are quoted in the document SRH 349, which appears earlier in this volume. The leller was hand-carried 10 Dewey by Colonel Carter W. Clarke, Deputy Chief of the Military Intelligence Service and former head of the Special Branch. This is Clarke's personal account of his unique mission. In the end, Dewey did not make any secrets public, but it took some persuading. STATEMENT FOR RECORD OF PARTICIPATION OF BRIG. GEN. CARTER W. CLARKE, GSC IN THE TRANSMIlTAL OF LElTERS FROM GEN. GEORGE C. MARSHALL TO GOY. THOMAS E. DEWEY THE LAlTER PART OF SEPTEMBER 1944

*** In the forenoon of Monday, 25 September, I was asked by Maj. Gen. Clayton Bissell , A. C. of S., G-2, if I was known personally to Gov. Thomas E. Dewey, Lt. Gen. Hugh A. Drum, or any of Gov. Dewey's entourage or close associates. [ stated that I did not know and, so far as I knew, was not known to any of these people. Gen. Bissell later informed me that I would be directed to take a letter to Gov. Dewey at Tulsa, Oklahoma, Gov. Dewey's next stopping point on his trip east from the West Coast. Later in the day, in the office of Col. Frank McCarthy, Secretary of the General Staff, I was given the letter to read, to check for accuracy of technical details. Later in the day I was informed by Gen. Bissell that I should travel in civilian clothes and, if possible, make my contact with Gov. Dewey unknown to any other person, but that under no circumstances was I to disclose to anyone except Gov. Dewey my mission or the contents of the letter, nor was I to give the letter to Gov. Dewey unless we were the only two present in the room at the time. I left the Pentagon Building early in the afternoon of Monday, 25 September, and went home in order to get my civilian clothes out of storage and have them pressed in preparation for the trip. After dinner I returned to the Pentagon Building, where I received orders from Major Earman at 1942 EWT, Monday, 25 September. Went to airport and got off ground at 2040 EWT. Landed at St. Louis for fuel at 2330 CWT. In air again 2400 CWT and landed in Tulsa at 0150 CWT. Slept in

172

SIGNALS INTELLIGENCE

Hotel de Gink. Up at 0700. Phoned Gen. Bissell at 0730. Went to Meyo Hotel, had breakfast and phoned Roger Randolph at 0820. Got haircut and shave and saw Roger Randolph at 0905. He contacted Mr. William Skelly, President of the Skelly Oil Company. Mr. Skelly was Gov. Dewey's campaign manager and was handling all arrangements. Went to Mr. Skelly's house and made arrangements to meet him in private hotel room in Tulsa Hotel. Met Mr. Skelly as agreed at 1600. He saw Gov. Dewey's Secretary who came down twice to insist that he know who I was and what I wanted. No one knew anything about the case except that I had a letter that had to be delivered to Gov. Dewey personally and alone. Finally he said, "You must write down the name of the man who has sent you, seal it and give it to Gov. Dewey." I did this, and after being gone 16 minutes, he returned and said Gov. Dewey would see me. I saw Gov. Dewey alone, delivered the letter and asked that I not read it aloud. Gov. Dewey opened the outer envelope and then said, "Well, Top Secret - that's really top isn't it?" He read the first two paragraphs and then stopped. He asked if I were a Regular Army officer. When [ said yes he asked if I would give him my word of honor that I had been sent by Gen. Marshall. I said I would. He then said he did not want his lips sealed on things that he already knew about Pearl Harbor, about facts already in his possession or about facts which might later come into his possession from other sources but which, if they were contained in Gen. Marshall's letter, could not be used because he had given his word on this letter, thereby sealing his lips. He said he would be glad to discuss this matter with me now. I told him I was merely a courier and was not authorized to enter into any discussion about what he knew about Pearl Harbor or about the contents of this letter. He then asked if I was authorized to say to him in the name of Gen. Marshall that if he read the letter through and then stated to me that he already had in his possession the identical information that was contained in the letter, that he would then be released from all obligations to keep silent. I said I had no such authority. He then said he could not conceive of Gen. Marshall and Adm. King being the only ones who knew about this letter. Furthermore he said he could not conceive of Gen. Marshall approaching an "opposition candidate" and making a proposition such as was apparently contained in that letter. He said "Marshall does not do things like that. I am confident that Franklin Roosevelt is behind this whole thing." All this time the letter was lying face down in his lap. He then picked it up and said "Let me read those first two paragraphs again." He started to read, then laid the letter down and said "I have not reread them because my eye caught the word "cryptograph." Now if this letter merely tells me that we were reading certain Japanese codes before Pearl Harbor and that at least two of them are still in current use, there is no point in my reading the letter because I already know that." He then said "That is the case and I know it, isn't it?" I said, "Governor, I am merely a courier in this case." He said, "Well I know it and Franklin Roosevelt knows all about it. He knew what was happening before Pearl Harbor and instead of being

THE POLITICS OF COM1NT

173

reelected he ought to be impeached." He then said "Would you like for me to phone Gen. Marshall and say to him what I have just said to you?" I said "Governor, this is a subject that should not, under any consideration whatsoever, be discussed or even mentioned over the telephone." He handed me back the letter saying, "[ shall be in Albany Thursday and I shall be glad to receive you or Gen. Marshall or anyone Gen. Marshall cares to send to discuss at length his cryptographic business or the whole Pearl Harbor mess. " He continued, "I can arrange it so the conference can be held in absolute secrecy." I then put the letter back into the envelopes, thanked him for his courtesy in giving me so much of his time, and departed. I then left the hotel, went to the field, phoned Gen. Bissell. Took off at 1815 CWT 26 September. I arrived at Washington airport at 2245 EWT 26 September. 1 phoned Gen. Bissell and asked if it was necessary for me to make my report that night. He said "No but be in the office early the next morning in order that we may report to the Chief of Staff." About 0740 on the morning of 27 September, accompanied by Gen. Bissell, I made oral report to Gen. Marshall as indicated above, leaving with him at that time Gen. Marshall 's letter which I had taken with me to Tulsa and brought back on my return trip. Later in the day on Wednesday, 27 September, 1 was informed by Gen. Bissell that on the following morning I was to go to Albany to contact Gov. Dewey and to give him another letter which had been prepared by Gen. Marshall. At 1008 on 28 September I took off from Washington airport and landed at Albany at 1205. Went by taxi direct to the State House. Arrived there at 1250. Contacted Gov. Dewey's office and was put in touch with Miss Ross, his private secretary. She asked who 1 was and what 1 wanted. I told her that 1 was same individual who had contacted Gov. Dewey last Tuesday in Tulsa and that I was there at Gov. Dewey's invitation with another letter and message for the Governor. She left to phone and in a few minutes came back saying that Gov. Dewey desired me to come right over to the Executive Mansion. I arrived there at 1330 and was taken directly to Gov. Dewey's reception room. Present were Gov. Dewey and a man introduced as Mr. Elliott Y. Bell , Superintendent of Banks, State of New York. 1 stated to Gov. Dewey that at Gen. Marshall's direction and in accordance with Gov. Dewey 's invitation extended in Tulsa, I was there to present him a letter and message from Gen. Marshall and to answer any questions of a technical nature that the Governor might care to ask after a reading of the letter. Gov. Dewey asked if it were the same letter 1 had in Tulsa. I said no but that it was on the same subject and that the meat of the letter was the same. He then asked if I still had instructions to see him alone. I replied that 1 had. He next asked if I could leave the letter with him. 1 said no that I was to return the letter to Gen. Marshall. He asked if I were authorized to discuss the contents of the letter with him. 1 stated that 1 could discuss the technical features of the case freely and could answer any technical questions on the subject that he might care to ask.

174

SIGNALS INTELLIGENCE

Gov. Dewey then said that, in view of the circumstances, his own personal position, in view of what he already knew, to say nothing about what he might later learn, and particularly in view of his trusteeship, he had decided that he could not see me or anyone else alone, that he would not read a letter which he could not keep and which he could not show to and discuss with Mr. Bell , nor would he enter into any further discussion without Mr. Bell being present. He said he could not afford to read a letter he could not keep as he might later be charged with reading a letter different from the one he had really read. He said I might be curious to know whether or not he had a recording device in the room . , said "No, Governor, , am not curious; I merely assume that you have one." He said "Well I haven ' t and I did not ever have one when I was District Attorney. I had one in my witness room, of course, but never in my own private office." He next said, "Now Colonel , Mr. Bell is an American citizen and is just as patriotic as any member of the Administration. He already knows all the facts that' know in regard to Pearl Harbor and I certainly intend that he know the contents of the letter if I read it. If' keep the letter' shall put it in my vault and no one but Mr. Bell and myself shall know of its existence or of its contents and we shall discuss it only between ourselves. Furthennore I do not see what Gen. Marshall is so exercised about. There are at least 12 Senators that I can name for you right now if you desire that know aU there is to be known about Pearl Harbor and about how we were reading certain Jap codes before Pearl Harbor and how it is claimed that we are still reading·two of these same codes. You know, Colonel, this code business is the worst kept secret in Washington, but I for one want to say to you that' do not believe any such thing as that to be a fact." I said, " Well , Governor, whether you believe it or not, it is a fact and' am willing to take an oath to that effect right here and now. Let me assure you that Gen. Marshall 's sole interest in this matter is to protect our most vital source of intelli gence and to save the lives of thousands of troops that are certain to be sacrificed if security on these ciphers is blown and the Japanese change them." I then said, "The re are man y other aspects of this case that I would like to tell you but' feel that you should read the letter first." He sat silent for about three minutes then said, " Why don 't you phone Gen. Marshall and tell him that J will not discuss the matter with you unless Mr. Bell is present, nor receive the letter unless I can let Mr. Bell read it and then' keep it T J said that I did not want to phone from the Executive Mansion but would go to a pay phone and call; that if J were unable to reach Gen . Marshall' would fly back to Washington at once, deliver Gov. Dewey 's message, and return to Albany tonight and be available to see him (Dewey) early tomorrow morning. He said. "Well' have a direct line in he re and I have a line that goes through the switchboard and there is no tap on either of them unless O 'Connor has just put one on. I have them checked several times a week so I feel that they are safe." , asked who O'Connor was and he said he was the Democratic leader here in Albany. I said, " Well' prefer not to have the phone used in thi s case." He said, " Oh hell , I'll

175

THE POLITICS OF COMINT

phone Marshall , I've talked to him before and this will be all right." He then picked up the phone and put in a call to Gen. Marshall. While waiting for the call, which took about seven minutes. Gov. Dewey said, "If, as you say, the Japs are still using two of their codes that they used before Pearl Harbor, why in hell haven 't they changed them , especially after what happened at Midway and the Coral Sea?" I explained the difference in the types of codes and ciphers used by the various Jap agencies, their di stribution problems: and gave the reason we think the Japs have not changed these codes. He then asked why Gen. Marshall was so anxious to stop him from talking when everyone else in Washington knows the story and is talking about it. Just then the call was completed and he talked to Gen. Marshall in re the letter and about Mr. Bell. He then gave the phone to me, and Gen. Marshall authorized me to give the letter to Gov. Dewey. to leave it with him and to discuss the case technically in the presence of Mr. Bell. As I was opening the envelope Mr. Bell sa id. "Colonel, hundreds of people know all about the Midway affair and how most of our other successes in the Pacific have been due to our reading Jap naval codes. Everyone who has ever been out there knows about it and talks freely about it. Why not long ago at a dinner where a large number of people were present I talked to a naval commander who had been out there and had participated in nearly every engagement we have had. He said that they always knew where the Jap ships were and that our people were told by radio where to station their own ships to meet the Japs, and that all this information came from reading Jap codes."

Gov. Dewey said " Yes and I have lots more information about the pre Pearl Harbor messages you were reading." He then took the letter and started to read. Mr. Bell said, "Tom , do you want me to read it over your shoulder?" Gov. Dewey said "No 1"11 hand it to you a page at a time." Gov. Dewey continued to read , then put the letter down and said, "Well I' ll be damned if I believe the Japs are still using those two codes." I again assured him that they were and that one of them was our life blood in intelligence. The letter was still in his lap when I said "Governor, I would like to repeat that I know that Gen. Marshall's sole interest in this case is to preserve the only worthwhile source of intelligence that this nation has. You do not realize it but the War Department has 10,200 people working in the Signal Security Agency; the Navy Depart ment has almost 6,000 in their Communications Annex, and both the Arnly and Navy each have several thousand in tactical field and fleet units engaged in signal intelligence work.

*** Gov. Dewey said "Yes, Gen. Marshall just said that we had over 10,000 on that work. "

I then stated that Churchill considered this hi s secret weapon and that it had really saved England. I described how Churchill felt about security, how the Navy

176

SIGNALS INTELLIGENCE

pri zed it so highly and how difficult it had been to break down Briti sh resistance because of American lack of security consciousness. 1 quoted to him Churchill's reported statement about protecting thi s source, how that in order to protect the source the Briti sh had time and again permitted convoys to be attacked rather than divert them from thei r course and thus blow security. Gov. Dewey then picked up the letter and read some more. Then he put it down and said "Is thi s true about the ass?" I said it was. He said " Well I don ' t feel towards the ass like most of the Army and Navy peop le 1 ha ve talked to. I have a lot of my men in it and they are all damn good, excellent fellows in fact, and they tell me that we have no really trained intelligence personnel in either the Arnly or Navy. They say that in both the ArnlY and Navy the officers are merely on temporary detail and that they then go to sea or back to a regiment or somewhere, and that most of the officers of both serv ices are woefully ignorant of world affairs. I suppose that ma y be the real reason for taking no action on the Pearl Harbo r warnings they had- they could not interpret the warning." Mr. Bell then said " Well it don 't account for them not doing anything about that message that gave them the warning that Japan was going to attack the United States-you know the one I mean- where they set the deadline for Kurusu for November 25 and then postponed it till November 29 saying to get done with their negotiations by that time or that things wou ld automati call y begin to happen. Anyone shou ld have known that that meant they were going to attack the United States. That was just the time limit on sailing dates. " Gov. Dewey said " Yes that 's right." Mr. Bell then said "Those messages you did not get till December 8 and after, why did it take so long for yo u to receive them?" I said "1 do not know what yo u are referring to, but before Pearl Harbor all our traffic intercepted on the West Coast, Hawaii and elsewhere was sent in by air mail which of course delayed it. " Gov. Dewey then read the letter through and handed the last page to Mr. Bell. After Mr. Bell had finished reading, he handed the letter back to Gov. Dewey. After a minute or so Gov. Dewey sa id " Well , except for the matter of the two codes and the ass, there is little in this letter that 1 did not already know. There is one point though, what in hell do Jap codes have to do with Eisenhower?" I exp lained the linkage with a general statement on the three types of ciphers used by G.A.F. , Abwehr, Clandesti ne, German Navy, Oshima 's visit s and interviews , etc. He seemed sat isfied. He then said "Colonel, I kno w a hell of a lot about thi s business. You probably know that some years ago when Stimson was Secretary of State he got very damn pure minded one day and stopped all such th ings as thi s. He afterwards denied that he had done it but said he was acting on orders from President Hoover. Well , as you know, the man who headed the cryptographic work of the Government was Yardley who was fired by Stimson. Yardley was a friend of mine in those days and I am the man who stopped Yard ley from publi shing hi s second book. I met him in a speakeasy one night and persuaded him aga inst publishing the

THE POLITICS OF CO MINT

177

book . He agreed provided that Mr. Stimson write him a leller of apology for the abusive manner in which Stimson had fired him . Stim son did, too. " I did not comment on thi s other than to say I was quite familiar with the Yardley case. I then asked if there were any questions the Governor or Mr. Bell would like to ask. Gov. Dewey said " Not just now but J would like to discuss this case with Mr. Bell. " r got up to go out of the room into the hall but Gov. Dewey said " No you stay here, we will go across the hall." I sat do wn and Mr. Bell said " Well , Governor, we have pledged ourselves not to discuss this lelter with anyone else but just between ourselves but what about Colonel Clarke-who is he going to discuss it with?" Gov. Dewey sa id, "Oh he will of course report this conversation to Gen. Marshall-is that not correct?" Before I could reply Mr. Bell said " Yes but with who else?" Gov. Dewey said " Anyone else?" I said, " I shall report to my immediat e supe rior, Major General C layton Bis se ll , A. C. of S. , G -2, War Department, and through him to Gen. Marshall and to no one else." They then got up and left the room , leav ing the leiter on a stand by Gov. Dewey 's chair. They were gone about 22 minutes . When they returned Gov. Dewey said "Let me look over this letter again and read carefully a couple of paragraphs." When he fini shed he said, " Well, Colonel, I do not believe that there are any questions r want to ask you nor do I care to have any discuss ion about the contents of the letter. Will you give me your full name, serial number, office location and phone number, and your residence address and phone number?" I did thi s and he thanked me and asked whe re r was staying. r to ld him r was Oying back to Washington tonight and was going direct to the airport. J then asked him if he had any message that he wanted me to take to Gen. Marshall. He said "No, no message." I thanked him for giving me so much of his time. He shook hands and said " Weill hope we meet aga in under more auspicious circumstances." I said goodbye to him and to Mr. Bell and left for the airport at 1610. Next I phoned Gen. Bi ssell and to plane. In air at 1645. I arrived at Washington Airport 1855 and phoned Gen. Bi ssell advising him of my return and asking if it was desired that I make report that night. He answered in the contrary but directed me to be in the office ea rl y the following morning in order that we might make report to the Chief of Staff. Again about 0740 on Friday, 29 September, accompanied by Gen. Bissell I made my second report to Gen. Marshall as indi cated above.

CHAPTER IX Tactical COMINT The Anny signal s intelligence effort was a divided one in World War II. At the national level, the Signal Security Agency provided hi gh-level cryptana lys is. The intercept arm of the agency was the 2d Signal Service Battalion, which maintained large fixed intercept stations in a number of locati ons in the Continental United States and around the globe. However, in the field , theater commanders directed a separate tactical effort through theater-level signal intelligence services, which controlled a multitude of signal radio intelligence compan ies and (in Europe) smaller signal service companies. By the end of the war, numbered Army Air Forces also had their own signal intelligence elements in the fonn of Army Air Forces radio squadrons, mobile. The diversity and scope of the tacti ca l effort is shown by the following excerpted documents: SRH 124, Operational History of the 849th Signal Inte lli gence Service; SRH 048 , Summation of Operational Activity, Signal Service Detachment "0"; SRH 042, Third Army Radio Inte lli gence History; and SRH 227, History of the 126th Si gnal Radio Intell igence Company. The excerpt from SRH 049 , Technical Intelligence Transmitted Directly to G-2 12th Anny Group, ETO From 14 August 1944 to May 1945 , contains one day's intelligence "take" from tactical COMINT efforts in Europe. It demonstrates what could be gleaned from radio direction-finding and the exploitation of low- and med ium -leve l Gennan ciphers.

180

SIGNALS INTELLIGENCE

From SRH 124. Excerprs from rhe official hisrory of rhe Army's signal in relligence service in rhe Medirerranean Th earer of Operarions. The accounr shows rhe slow evolurion of Army signals inrelligence organizarion in rhe field as rhe resulr of lessons learned in acrual com bar operarions. As a marler of inreresr, the 849rh Signallnrelligence Service was the only World War /I rhearer signal inrelligence service ro be assigned a numerical designarion.

OPERATIONAL HISTORY of the 849TH SIGNAL INTELLIGENCE SERVI CE Medite rranean Theater of Operations United States Arm y

*** PART I: BEGINNINGS

GENERAL I. It was ori ginall y planned that all U.S. signal intelli gence activities at AFHQ would be handled by a unit entitled Signal Intelli gence Service, AFHQ, the commanding officer of which would be a me mber of the Chief Signal Officer's Staff. A TIO and an SLOE were submitted to the War Department on thi s basis. These were approved, but the unit was redesignated as the 849th Si gnal Intelligence Service, and it was decided that the unit would not constitute the Signal Intellige nce Divi sion of the Signal Section, but would work under that division in a separate location of its own at some di stance from headquarters. Since the re was no basis of ex perience in contemporary warfare on which to build, the original TIO mere ly represented a careful guess as to what the function of such an organization in an active theate r of operations should be, and what personnel would be necessary to perform them. The organization, being unique, was not referred to in any T / BA: and the Special Li st of Equipment, which was its only basis for drawing suppli es, would have been markedl y inadequate in any case. The fact that it was not modi fi ed in any way to COnf0n11 to the change from a staff section to a fi eld unit multi plied thi s inadequacy. As a result , the unit set up in an isolated ravine in North Africa, without a te lephone, a foot of fi eld wire, a radio set, or a power unit , to mention only a few of the more obvious necessities. Many of these items would ha ve been provided automatically for a staff section , but to procure (and keep) them for a field unit, entitled only to TIE and T/BA allowances has invol ved a com-

TACTICAL COMINT

18 1

plicated and continual struggle. 2. In order to take advantage of greater British experie nce in inte lligence operati ons, the 17 office rs, and 90 EM ca ll ed for in the Cryptanalyti c and Radio Intelli gence Sections of the Inte lli gence Branch were activated at FOri Dix o n 8 November 1942 as Signal Intelligence Detachment 925 1-A, under the command of Capl. [text wi thheld] .... The Detachment left for England on 24 November. It was onl y upon their arrival in Africa, three months later, that the members of this group learned that they were eanllarked fo r the 849th SIS, of which they had never heard. On 2 December 1942 the 849th SIS was form ally activated at Fort Devens, with Capl. [text wi thheld! in command. Assigned to it at this time we re 16 oflicers and 102 EM, whic h, with the personnel of Detachme nt 925 1-A made up the exact number ca lled for on the T/O. Captain [tex t withheld] was under the impression that the personnel of Detachment 9251-A had also been assigned, and no evidence was available by whic h thi s impression could be corrected . The unit was therefore apparently up to fu ll stre ngth ; but the C.O. was junior to Capt. [text withheld] who had originally been intended merely as a sect ion head . In general , technical qualifications and relati ve rank were so poorl y adjusted that it was clear that one or the othe r would have to be di sregarded unless additional offi cers were brought in at the top. The fact that the organizati on, including the detachment in England , contained only six officers above the grade of second lie utenant and none above the grade of captain , while the T/O called for a colonel. two lie ute nant colonels. and seven majors suggested that this was very likely to happe n. a. It did happen on 14 Ja nuary 1943, when five officers from SIS , Western Task Force and three from SIS , AFHQ were assigned by NATOUSA orders to the 849th SIS , which embarked from NYPE the same day. This gro up included Major !t ex t w ithh e ld] Commanding, and Cap l. [t ex t w ithh e ld] se lec ted as O IC. Intelligence Branch. Twenty seven e nli sted men from the same sources were also assigned to the 849th SIS, a site was found at Hammam Melouane. and operations were begun on such a limited sca le as was possible. On I February the main body of the 849th personnel arrived. and Capl. [text withheld] was designated O IC, Security Branc h. On 20 February, Signal Inte lli gence Detachme nt 925 1-A, less three officers and twelve men left in England for furth er training, arrived and was assigned to the 849th. The organization was now conside rably above its authorized strength. particularly in orticers; but in the midst of the gene ral confusion thi s fact escaped unfavorable comment. The key position had been fi lled, the general shakedown process was well under way. and the organization was ready to begin full scale ope ratio n under the supe rvision of Colonel Harold G. Hayes, SIS officer, AFHQ. It must be confessed that the question of wha t these operations were to be still needed a good deal of clarification. b. Three major po licy proble ms had to be faced : ( I) Coordination with the [text withheld! whose intelligence and securit y organizat ion does not parallel our own.

182

SIGNALS [NTELLIGENCE

(2) Determination of the levels at which various SIS activities were to be carried on and the inter-echelon coordination of these activities. (3) Adaptation of radio intelligence organizations to the intelligence and security needs of the theater. In addition there were the continually developing problems of procuring and training personnel , procuring, modifying, and improvi sing equipment, selecting specific commitments and devising techniques to meet them, etc. 3. After these three problems and those of supply had been solved to a degree which permitted operations to begin, another and more difficult problem became apparent as soon as the 849th SIS began to serve other headquarters. This was the non-acceptance by many commanding and staff officers of the benefits, intangible tho many were, resulting from an adequate signal intelligence program. a. The intelligence Branch met this problem in the reluctance with which American tactical staffs first allowed Detachments of the Branch to operate in combat areas, where tactical information available to intercept and cryptanalysis was most abundant. It was believed that such an apprehension for the safety and security of signal intelligence operations was founded on a general ignorance within the then new American tactical staffs of the role which signal intelligence could play in combat operations and an unfortunate misconception in the same quarters of how the exploitation of Gennan Army low-security radio traffic could best be accomplished. American tactical commands, unfamiliar with signal intelligence services, were not anxious for the information which signal intelligence could supply, and were at first skeptical of the validity of some of the infonnation produced by the Detachments. FurthemlOre numerous serious security violations were at first made by tactical commands in handling signal intelligence infonnation, illustrating the general lack of appreciation for signal intelligence methods in tactical commands. b. Unlike the ltext withheld] in North Africa, whose staffs had become thoroughly familiar with tactical signal intelligence operations as a result of the desert campaigns, the new American staffs in Algeria and Tunisia had to be indoctrinated with the manner and value of signal intelligence work. The critical period of "selling" signal intelligence to American staffs as an integral intelligence function in forward areas occurred during the invasion of Sicily. Detachments "A " and

"E" of the intelligence Branch were originally scheduled to go ashore in Sicily with the Headquarters Seventh Army and the reinforced 3rd Division (VI Corps), Detachment "A" to land as soon as a beachhead of 10 square miles had been secured, Detachment "E" to land within a few days after D-Day. Concern for the security of signal intelligence personnel and operations resulted in Seventh Army Headquarters holding the Detachments in North Africa until 7 August (D + 30), and the units did not land in Sicily until 10 August , one week before the end of the Sicilian campaign. With such a short period remaining for operations before the

TACTICAL COMINT

183

conclusion of the campaign the achievements of American signal intelligence units in Sicily were obviously not impressive. The Detachments furnished valuable tactical information throughout the short pe riod; but their g reatest contribution, which would have been possible in the early days of the campaign, when enemy strengths and disposi tions were largely undetermined , was lost to the Allied effort. c. Gradually, however, American tactical commands developed g reate r appreciation for intercept intelligence and the problems involved inproducing such intelligence; and in time sig nal intelligence was conside red at all times to be a dependable source of inte lligence and under some conditions to be the best source of infornlation concerning enem y tactical forces available to American Army and Air force tactica l commands. That Detachme nts were in time allowed to operate in forward combat areas where tactical info rnlation was best intercepted is demonstrated by the fact that two Intelligence Branch Detac hme nts landed at Anzio on DDay and were scheduled to land in Southern France on D-Day but were he ld up until approximately D + 5 because of confusion in shipping schedules. d. Early, effec ti ve operation s by the Security Branch we re hinde red through a failure of staff and unit officers to appreciate or even reali ze that all s ignal sec urit y efforts were directed at protecting them and their units from e nemy ac tion . It was thi s factor whic h made it impossible to correct quickly the uniformly lo w state of s igna l securit y training with which American troops entered No rth Africa. The general attitude toward s signal securit y was that it consisted of a set of c umbersome regulations, and the reaction varied according to te mperame nt. Many officers we re frankly hostile, considering all SIS pe rsonne l as parasiti ca l non-combatants consp iring to compli ca te the jobs of honest fighting men. Others, havi ng developed a philosophical resignation toward all regulation s, were almost pathetically eager to comply with whatever was publi shed ; .. . .ln operations, the first reaction resulted in so man y messages passed in c lear and others so care less ly e nciphe red that it looked like a conspiracy to ignore as many sec urity provi sions as possible and in the second , resulted in an attitude whi ch sa id in effect " put it in some sort of code, that is what the book says". The idea that the e ne my mi ght use inte rce pted informati on with unfortunate res ults to themselves was almost en tirely abse nt. It was only afte r the loss of much valuable time and intense missionary wor k that an understanding of the purpose of signal sec urit y was ac hieved . e. Time and furt he r contact with the enemy brought changes in these atti tudes. By the close of host ilities in the MTO the importance of s ig nal intelligence was well accepted as an cssent ial part of a complete military effort and the 849th was being called upon to perform many functions in addition to it s primary mi ss ion. That a ll as pects of signa l intelligence activ iti es were eve ntuall y appreciated in Mediterranean Theater of Operations can best be illustrated by the 70 letters of commendation and 42 awards received by approximate ly 400 ind ividuals of the 849th S IS ....

184

SIGNALS INTELLIGENC E

4. Atlached ... is a table of personnel origina ll y allotled to the 849th S IS. It is shown here wi th the caution that it proved inadequate for signal intelligence activities in MTO and that at best can on ly be uscd as a guide in planning similar units since SIS functions appear to vary greatly from theater to theater. The scope of the 849th early increased beyond the bounds of the TIO. The personnel problem however was alleviated to a large extent by the assignment of an AFHQ Signal Intel ligence pool of 27 officers and 284 en listed men and an AFHQ Cryptographic pool of 29 officers and 323 enl isted men. While this appears to be a tremendous increase ovcr the TIO, the 849th also took over the mission o f those two pools. The additiona l personnel permitted the headq uarte rs section to be a ugmented where needed and permitted a shifti ng of officers and men to meet the ever chang ing operational demands, all of which great ly increased the effectiveness of the 849th SIS. PART II : INTELLIGENCE BRANCH GENERAL FUNCTIONS

I. Introduction. a. The basic duties of the Intelligence Branch were clear in principle from the date of activation of the 849th S IS: to derive intelligence from enemy radio transmissions. At that date, however, the multiple forms which Intelligence Branch functions wou ld later assume and the many problems which were to be encountered in fulfilling those duties cou ld not be foreseen. b. Before the end of hostilities in Europe the Inte lligence Branch wo uld develop from a handful of personnel to a complex organization whose activit ies included: three Detachments fulfilling signa l intelligence commitments at U.S. Corps (II, IV, VI) Headquarters, involving the interception and ana lysis of German Army low-security radio traffic: two similar Detachments fulfilling simi lar commitments at U.S. Anny (Fifth, Seventh) Headquaners; one Detachment analyzing German Army medium-security traffic: two Detachments providing the Mediterranean Allied Air Forces with information derived from the analysis of German Air Force low-security and plain-language (voice) radio transmissions: one Detachment analyzing enciphered European weather broadcasts cmanating from within enemy-held territories for the 12th and 15th U.S. Air Forces: one Detachment analyzing [text withheld] radio traffic and later providing the Office of Strategic Services with information from a study of clandestine radio traffic of enemy agents in Allied controlled territory: one Detachment providing the Civil Censorship Group of the Allied Control Commission with information from the analysis of secret inks and other types of unauthorized correspondcnce; and a Headquarters whose operations included, in addition to routine administrative function s, a continuous supply of technical intelligence and information to the Detachments in the field. research into thc German communications systems, solu-

TACTICAL COMINT

185

tion of German Army medium-security traffic, a Communication Section to provide all communications between the Headquarters 849th SIS and its Detachments and other Headquarters, and a Laboratory Section for reproducing important captured documents and photographing material s of all kinds. c. When the Intelligence Branch began operations in North Africa during the winter of 1942-43 radio intercept and the cryptanalysis of enemy low- and medium-security radio traffic within an active theater of operations constituted an unexplored field for the new United States Army. Consequently, the growth of the Intelligence Branch was all ended by many restrictive elements. [text withheld] It had been arranged to have several Signal Radio Intelligence Companies provide all American radio intercept facilities required within the theater, operating in conjunction with, but independent of, Intelligence Branch cryptanalytic Detachments; and it was not long before such an arrangement presented a number of shortcomings. Inadequacies in the 849th SIS T/O presented additional difficulties to the development of the Intelligence Branch; and the invariable insufficiency of training of all newly-arrived personnel from the United States and London was another stumbling block to the proper expansion of the Intelligence Branch. Finally, the problem of equipping each newly organized detachment was complicated by grave inadequacies of the TIE of the 849th SIS. 2. Cryptanalysis. a. In order to limit 10 a reasonable ex tent the scope of work to be performed by the Intelligence Branch, consideration being given to the number of personnel available and to the ir state of training, it was planned originally that the work of the Branch would involve at first only the study of German Air Force lowsec urit y radio traffic. Such traffic consisted of radio activity of the German Air Forces operating throughout the western and central Mediterranean area, involving transmissions of both enemy aircraft and ground radio stations. [text withheld] With thi s type of traffic in mind the first operational unit created within the Intelligence Branch, eventually called Detachment "F", was formed in February 1943. Personnel of this Detachment underwent preliminary training [text withheld] near Algiers, and by the summer of 1943 were ready and able to assume an independent signal intelligence operational comm itment. b. After the personnel of Detachment "FH had commenced their training and additional personnel were available to the Intelligence Branch consideration was given to another type of enemy radio traffic, Gennan ArnlY low-security traffic, most of which emanated at regimental level and below. This traffic consisted of Itext withheld] jargon codes, simple substitution ciphers and transposition systems, and was fairly easily broken also by elementary cryptana lysis when sufficient depth was encountered. At the end of February a small group of personnel was allached to [text withheld I United States II Corps in Tunisia to learn the techniques of analyzing such German Army low-sec urity traffic. These people fornled the basis of what lat er became Detachment "A". As time wcnt on additional

186

SIGNALS INTELLIGENCE

detachments were form ed to pe rform the same type of cryptanalytic function s as Detac hme nt " A", and eventually there were five detachme nt s ex ploiting such German Arm y low-sec urity traffic : Detachment "A" at Headquarte rs Fifth Army, Detachment "B" at Headquarters Seventh Army, Detachme nt " E" at Headquarters U.S. VI Corps, Detachment " H" at Headquarters U.S. 11 Corps, and the 39 16th Signal Serv ice Company at Headquarte rs U.S. IV Corps. c. During the summer of 1943 it became apparent that an additional cryptanalytic task could be undertake n by the Intelligence Branch. The Amly Air Forces were in need of the infomlation which could be derived from enciphered European meteorological radio transm iss ions g iving vauable weather data over enemy-held te rritories. Work on this mate rial was begun in July 1943 with the formation of De tachme nt "G". Inte mational Business Machine equipme nt was obtained; tcchniques for breaking the traffic were evo lved; and the latest meteorolog ica l observations for e nemy he ld te rritory were made available to tactica l and strateg ic air forces in the Mediterranean area. d. In July 1943 a furthe r development occurred when it was found possi ble to allempt the so lution of German Army medium -sec urity doubl y-e nc iphe red Playfair traffic. Pe rsonnel for this task, who had been trained in the soluti on of German Air Force medium -security traffic in London, arrived in the Mediterranean Theater in early Jul y. These people, who became the nucle us of Detachment " B" and later of the Solution Section at the Headquarters, Intelligence Branch, e njoyed considerable success in their solution operations whenever traffic in suffi cient depth was available. The greatest difficulty e ncountered in dealing with such traffi c, emanating usuall y from enemy Divi sional networks, was in finding suitable sites for its interception. e. The value of German Air Force voice traffi c on Very High Frequencies (38-42 mcs), mostly plain language, had long been recognized, and in August 1943 a detachment of German ling ui sts was fonn ed to explo it thi s source of intel li gence, eventuall y becoming De tac hment " D". Pe rsonnel were assigned aboa rd the USS Ancon, and the gro up provided important intelli gence of German Air Force activity during the land ings at Salemo. Later the Detachment operated in the vicinity o f Naples, on the Anzio beachhead, during the invasion of Southern France , a nd no rthward wi th th e A lli ed force s throughout France a nd into Germany, worki ng wi th Allied Tactical Air Com mands and Fig hte r Wings. The same type of traffi c was also exploited during more than a year's period by two groups of three e nlisted men of Detachment " D" aboard Un ited States destroyers escorting Allied convoys throughout the Mediterranean area. After Janua ry 1944 German linguists were used regularly in all Inte lli gence Branch Detachments serving Corps Headquarters to intercept German Army voice traffic on such Very High Freq uencies (32-38 mcs).

***

TACTICAL COMINT

187

4. Laboratory Facilities. a. Communications Intelligence. A small section had originally been provided at the Headquarte rs, Intelligence Branch for the collection and study of captured enemy communications equipment. It was soon felt, however, that the section was unable to carryon its prescribed duties successfull y as a subord inate part of the Intelligence Branch, and in June, 1944 the section was activated as the Enemy Equipment Intelligence Service. Having its own T/O and E, the new unit was assigned to the 849th SIS and removed from the control of the Intelligence Branch .... b. Laboratory Section-Photographic and Chemical. Photographic Facili ties provided by the original 849th SIS T/O and E proved inadequate to the tasks which later developed, and constant additions in personnel and equipment were necessary to meet constantly greate r requirements. In addition to providing photographic facilities for the 849th SIS the Laboratory Section was responsible for maintaining a laboratory to make c hemica l anal ys is of civi lian and prisoner-of-war mail for secret ink and other unauthorized means of communication. Original estimates of personnel, equipment, a nd s uppli es necessa ry for the C he mical Subsection were similarl y disproportionate to actual requirements, and a process of continuous g ro wth was characteristic of that subsection also. c . Communi ca tion s Sec tion-From a small g roup of personnel who installed telephone and lighting facilities at Hammam Melougne in February 1943 was grad ua ll y evo lved a Co mmunication s Section. Although pa rt of th e Intelligence Branch the section provided communicat ions fac ilities for the e ntire Headq uarters, Intelligence Branch. [n time the section expanded in size and operations until it maintained radio a nd teletype links with most field units. MAJOR PROBLEMS ENCOUNTERED

*** a. At the opening of operations in the Mediterranean Theater [text withheld] already had a numbe r of signal inte lli gence units in the field properly trained for the inte rception and so lution of German Army and Air Force low-security radio traffic. Plans for utilizing U.S. signal intelligence personnel had not been properly integrated with the British before the invasion of North Africa; but shortl y after the f0I111ation of the 849th SIS and its arrival in North Africa [text withheld] it was recognized that American units wo uld in time be able to contribute to the North African signal intelligence effort.

***

2. U.S. Signal Radio Intelligence Companies. a. Four U.S. Signal Radio Inte lli gence Com panies-the 11 7 th , 122nd, 123rd , and 128th- together with a number of intercept ope rators later assigned to

188

SIGNALS INTELLIGENCE

the 849th SIS itself provided intercept facilities for the several detachments of the Intelligence Branch. In a number of cases the SRI Companies were themselves split into detachments to provide facilities for new Intelligence Branch detachments fonned 10 meet new commitments. All four SRI Companies were originally assigned to Allied Force Headquarters (AFHQ), and their operational activities were then directed by the American Signal Intelligence Officer of AFHQ. b. Of these four SRI Companies the 122nd and 128th arrived in North Africa with the invasion forces. The operators were inexperienced with the new types of traffic they were to intcrcept; and until late 1944 there was an insufficient number of trained operators to meet the minimum requirements of the indi vidual Intelligence Branch detachments and monitoring tasks which were assigned by the Monitoring Section of the Security Branch, 849th SIS. c. Early in 1943 steps were taken to remedy the shortage of intercept operators by securing the I I 7th and 123rd SRI companies to supplement the first two. Before these units arrived in North Africa several important months had passed, and more months of train ing were required to bring the operationa l efficiency of all units to a point where specific commitments were possible.

d. A directive from the American Signal Intelligence Officer, AFHQ, made clear the fact that the Intelligence Branch Detachments working in conjunction with the SRI Companies were to direct the employment of sets and assign specific intercept missions. Experience proved, however, that although relations between the SRI Companies and the Detachments of the Intelli gence Branch were on the whole satisfactory the degree of cooperation and efficiency obtained often depended on purely personal relationships between the officers and men of the units in question and not on any clearly defined policy of command. The separation of intelligence and intercept sections, although possible in theory, in some cases produced situations in which questions of policy and operational control were in open conflict. The fact that the Intelligence Branch Detachments had to be attached to the SRI Compan ies for rations, quarters, and adm ini stration produced ot he r unsatisfactOlY situations that hindered proper cooperation between the two types of units. e. A step toward remedying the s ituation was taken by the Army Air Forces after the assignment of the 122nd and l23rd SRI Companies 10 the 12th Air Force in November 1943. A plan was studied and later approved 10 provide a se lfconta ined signal intelligence unit comprising both inte lligence and intercept personnel. On the basis of such reorganization the 123rd SRI Company was renamed the 9th AAF Radio Squadron Mobile (G) in March 1944; after a long delay intelligence personnel were obtained for this new Squadron from the U.S. , and the Squadron began to function as a com plete unit in December 1944. The 122nd SRI Company was never reorganized, and con tinued to operate in conj unction wi th the Inte lli gence Branch Detachment "G" until June 1945. f. A simil ar step in the right direction within the Army Ground Forces occu rred when TIO's and TIE's were approved by the War Department in January

TACTICA L COMI NT

189

1945 for Corps Signal Service Companies, prov iding for self-suffi cient organi zations which contained both intercept and inte lli gence pe rsonnel. Detachme nt " H" and a detachment of the 128th SRI Company whi ch had been ope rating in conjunction with Detac hme nt " H" were combined to fornl the nucleus of one such Signal Service Company (39 15th SS Co) in February 1945 .... g. Anothe r critical handicap encountered as long as the SRI Companies and Intelligence Branc h Detac hments operated independe ntl y invo lved a theater policy concern ing di sseminati on of signal intelligence infollnation. Accord ing to theate r regulations pe rsonnel of SRI Companies we re not all owed access to techni ca l signal intelligence infonnation and proced ures that formed the bas is for inte rcept set assignme nt. Conseq ue ntl y SRI personne l were g iven o nl y speci fi c freque ncy and callsign miss ions dail y, and we re not entitled to know the basis fo r such missions. Such an arrangement required addit ional signal intelligence personnel on duty in inte rcept rooms supe rvisi ng inte rcept acti vi ties to guarantee proper coverage. And in spi te of such efforts many va luable minutes of intercept operators' time we re often wasted by copying worthless traffic before being noticed by signal intelligence personnel on dut y in the inte rcept rooms. Both intelligence and intercept pe rsonnel resented the restriction; but as long as they ex isted unde r separate units nothing could be done. Inte rcept pe rsonnel were espec ially resentful of the Fact that they we re de nied access to infornlation which wo uld increase the effi ciency o f their wo rk. The situation was corrected onl y whe n the two types of personnel we re pl aced unde r o ne co mm a nd , as in the case of the 39 15 th Sig na l Serv ice Company; fo r then both inte rcept a nd inte lligence personnel were allowed complete access to all pertine nt signal intelli gence data. 3. Proc urement and Training o f Personnel. a. When the 849th SIS was ac ti vated it was expected that the Intelligence Branc h wo uld consist onl y of Signal Intelligence Detachme nt 925 1-A; but, in fac t, it was subsequentl y composed o f several he te rogeneous groups, each trained fo r d ifferent tasks. SID 925 1-A. after preliminary training in cryptanal ytic methods for a short period at Vint Hill Station, USA , e mbarked for SIS ETOUSA in November 1942 fo r practical experie nce in processing tra ffi c from both Ge rm an Arm y and Air Force units. Of thi s group 14 o ffi cers and 77 e nlisted me n we re sent to the 849th SIS in Februa ry 1943. leav ing 3 offi cers and 13 enli sted me n in Londo n (eventua ll y to become nucleus o f So luti on Section, Headqu arte rs Inte lli gence Branc h) to deal operationall y with German Air Force medium-securit y traffic. b. During Jan ua ry 1943 the inte ll igence membe rs o f the Western Task Force SIS Section consisting of 2 officers and 6 enli sted men hav ing comple ted their d uties in Morocco, had set up their headquarters at Hammam Melouane, near Alg ie rs. One officer and one enli sted man from AFHQ jo ined them the re, and with the arri val of the London gro up in February pla ns for training and ope rations in the theater we re put into effect. c. The largest single inc rement o f pe rsonne l became avai la ble whe n the

190

SIGNALS INTELLIGENCE

T/O for the Signal Intelligence Detachment. NATOUSA, was assigned to the 849th SIS in June 1943. Thanks to thi s pool additional personnel we re subseq uentl y available to fulfill the demands of new commitments as they arose. d. Constant additions to the Intelligence Branch were, however, required since that time. Throughout the hi story of the Branch it was a never-e nding strugg le to get adequat e ly trained offi cers and e nli s ted me n to perform the tasks required of the Branch. The process of obtaining personnel from the USA was almost always a slow and unsatisfactory one. Frequently men arrivi ng in the theater proved to be totall y unfit for the duties fo r which they were requested and hence co uld not be used. Fo r exa mple , a request was sent to Washingto n in February 1943 for immed iate Shipme nt of 10 offi cers trained in elementary cryptanalysis a nd , more import a nt, with flu e nt knowledge of spoke n and written German.The 10 offi cers arrived at the 849th SIS at the end of May 1943, afte r the close of the Tunisian Campaign; and of the ten officers only two had any knowledge of the Gennan language. Another example of the inadequate training ex pe ri enced by many people who were assigned to the 849th SIS was the case of pe rsonnel of the Weste rn Task Force. who had received only 10 days ' instruction in the problems of radi o intelli gence pri or to the invasion of North Africa and who had been given no idea of the code and c iphe r mate rial they might be called on to deal with. Consequent ly training of personnel was the most important single problem that faced the Intelligence Branch. No single increment arri ving at the 849t h SIS had had a sufficientl y long training period unde r anything approaching operational conditions before entering the Mediterranean Theater. Pe rsonnel trained at SIS ETOUSA for a month or two bene fitted greatly from instructions given the re [text withheld I but the burden of forming detachments into units capable of assuming ac tual commitme nts rested e ntire ly on the 849th SIS. Expe rie nce proved that most detachmcnts required from three to five months ' experience unde r field conditions to be able to unde rtake successfully any operational commitme nt. e. It should be noted that too few seni or officers with adequate training were available to the Intelligence Branch during the early days of the 849th SIS. The majority of officers compe tent to handle the technical problems of the va ri ous detachment s we re in the grade of Lieutenant and were conseque ntl y placed in a difficult position as regards efficient liai son with the operational staff's to which they were furni shing intelligence. Over a period of two years the situati on gradual ly adjusted itself: but it is felt that had the initial handicap not ex isted ex isted there wo uld have been considerabl y smaller loss of valuable time and. perhaps, li vcs. 4. Organization and Equipment of Field Detachments. a. As has been stated above, the original T/O and T/E of the 849th SIS were manifestly inadequate to meet the great expansion of the organization in the past t wo and o ne ha lf years. This prob lem was espec iall y ev id e nt when Dctac hme nts we re formed for serv ice in the field. The only poss ibl e way for detachments to start operations was by attaching them to other se lf-contained orga-

191

TACTICAL COMINT

nizations, such as the SRI Companies, for rations, quarters, and administration. Because of the steady development of the Branch; grades and ratings were not allotted to the Detachments on any fixed basis; and equipment for the Detachments consisted of such supplies as were available to the 849th SIS as a whole and could be spared from other duties. Such supplies and equipment as were available were issued on memorandum receipts to the Detachments. Allocation of grades, ratings, supplies, and equipment to the Detachments was made by the Officer in Charge of the Intelligence Branch on the basi s of what was hoped to be a fair distribution among the Intelligence Branch, Security Branch, and the 849th overhead. b. Such conditions were obviously unsatisfactory and unconducive to a proper spirit of morale among the Detachments. No definite g rades and ratings could be counted upon with certainty within the Detachments , and Officers in Charge of the Detachments experienced difficulties from being in a subordinate administrative position to the Commanding Officers of the SRI Companies. c. Basic TIO's for Signal Intelligence Service Detachments types A, B, C, 0 , and E were in existence si nce December 1943, but Theater regulations did not permit activation of any such detachments until June 1944, when an "A" type and a "B" type unit were activated as the 3200th and 320 I st SIS Detachments assigned to the 849h SIS .... The personnel for these new Detachments were obtained from the Intelligence Branch Detachments "A" and "E" respectively, and some gains in operational efficiency were noted from the reorganization. Great improvement was particularly noted with respect to grades and ratings, for with their own TIO's at their disposal the officers commanding the new detachments were assured definite grades and ratings and cou ld handle their own promotional problems. On the other hand such reorganization did not so lve the basic problem of the relation ship between the Detachments and the SRI Companies, for there was no provision within these new Detachments for intercept personnel; and consequently operations with independent SRI Companies were still necessary. d. The most sati sfactory solution was reached when the TIO for Corps Signal Service Compan ies, approved by the War Department in Jan uary 1945, became available in the Mediterranean Theater. Two such units were activated as

the 3915t h and 3916th Signal Service Companies, and were assigned to the 849th S IS .... The new companies, calling for both intelligence and intercept personnel under a single command, eliminated entirely the problems of grades and ratings, supplies and equipment , and operational control. e. In spite of the advances made by the activation of such units three Detachments of the Intelligence Branch- Detachments "0", "F", and "G"--could not be reorganized, and operated throughout their existence without benefit of special T/O's. TIE's, or administrative independence of the SRI Companies with whom they worked.

***

192

SIGNALS INTELLIGENCE

From SRH 048 . Once American forces had estahlished themsel ves in France, Signal Security Detachment "0" served as the central commandand- control element for all signals intelligence operations conducted by General Omar Bradley's 12th Army Group. SUMMARY OF OPERATIONAL ACTIVITY OF SIGNAL SECURITY DETACHMENT "0", COVERING THE PERIOD I SEPTEMBER 1944 TO I APRIL 1945 I. PURPOSE a. Thi s report is intended to summarize briefly the signal intelligence miss ion assigned to Signal Security Detachment " D" and actions taken by that Detachment in the accompli shment of that mi ss ion. Although the Detachment began field operations as a unit in the middle of August 1944, I Septembe r was chosen as the starting point of this report since effective contacts with G-2 Section, 12th Army Group Tactical Headquarters, and with the signal radio intelligence units in the field were firmly established by that date. 2. M1SSION OF THE DETACHMENT a. The mi ssion assigned to the De tac hme nt was the proc urement of Si gnal Intelligence rrom the solution of enemy (German) medium and low grade tactical codes and ciphers for the AC of S, G-2, 12th Ann y Group, the coordination and techni cal supervi sion of the Si gnal Radio Intelligence Companies and Si gnal Service Companies operating under units assigned to 12th Anny Group, technical liai son with adjacent Anny Group Signal Inte lligence units and with those of highe r headquarters, and the coverage of clandestine radio traffic for study by the Radio Security Se rvice (British) and as required by the AC of S, G-2, 12th Amly Group. The latte r miss ion, being somewhat separate and distinct from the provi sion of signal intelli gence, will not be covered in this report. b. The mission of the Detachment may be conveniently divided into four di stinct responsibilities: ( I) To Signal Intelligence Divi sion, Hq ETOUSA. (2) To G-2 Section , Tactica l Headquarters, 12th AmlY Group. (3) To Signal Radio Intelligen ce and Signal Se rvi ce Companies, 12 th Army Group. (4) To flanking Almy Groups (6th and 2 1). c. The manner in whi ch the mi ssion was carried out , will be described in the secti ons of the report which follow. 3. SIGNAL INTELLIGENCE DIVISION , HEA DQUARTERS ETOUSA a. Signal Security Detac hme nt " D" was charged wi th the respon sibilit y of e nforc ing the general sig nal intelligence policies of the War Depa rtm e nt. Hq ETOUSA, and the 12th Anny Group, among the Signal Radio Intelligence units or 12th Arm y Group. b. In addition, the detachment provided the higher headquarte rs with the raw

193

TACTICAL CO MINT

material and finished productions of all signal radio intelligence organizations of 12th Army Group. (I) All signal intelligence and technical data were forwarded to the Signal Intelligence Divi sion, OCSigO, Hq ETOUSA. This commitment included transmission to that office of all technical reports published by the Intelligence Branch of the Detachment. ... These reports were also disseminated to the Mediterranean Theater of Operations, to British signal intelligence agencies in the United Kingdom and to the War Department , Washington , D.C. , through the Signal Intelligence Division, Hq ETOUSA. (2) Radiograms and tel etypc messages containing current information were prepared and transmitted at least twice daily. These messages reported all data collected by the Detachment in the course of its daily operations. (3) Captured documents, received from signal intelligence units of 12th Army Group were processed, studies published, and copies forwarded to Signal Inte lligence Divi sion. Hq ETOUSA. for permanent files. 4. G-2 SECTION , TACTICAL HEADQUARTERS , 12TH ARMY GROUP a. Detachment "D" was charged with the respons ibility of passing to the Assistant Chief of Staff, G-2, 12th Army Group, signal intelligence procured from the analysis of enemy low grade and medium grade (THUMB, PEARL and ClRO PEARL) tactical traffic. This was accomplished in two ways. (I) Direct telephone communication was available until 4 March 1945 when, due to the tactical situation , the Detachment operated too far in advance of Tactical Headq uarters to allow utilization of metallic telephone lines. However, radio or teletype (or both) links maintained constant communication from that date onwards. The appointment on 3 February 1945 of a liai son officer from the Detachment to G-2 Section also aided in the smooth flow and interpretation of s ignal intelligence. (2) All pertinent technical reports published by the Detachment were transmitted daily to G-2 Section. These reports supplemented and enlarged upon the infom13tion already passed to G-2 Section by communications means described in paragraph a( I ) above. b. A comp lete record of all signa l intelligence. of immediate value only, tran smitted to G-2 Section by telephone, radio or teletype has been maintained by the Inte lligence Branch of Detachment " D", but thi s record is too bulky for reproduction and inclusion in this report. However. for the sake of completeness, an estimate of the quantity of timely information reported during the period under review may be derived from the following figures:

CIRO PEARL PEARL THUMB

No. of Items 1422 632 162

194

SIGNALS INTELLIGENCE

c. It is not poss ible to eva luate the signal intellige nce made available to G-2 Section , either in the form of immediately applicable infornlation (i.e., location of Division or Anny ammunition, supply and fuel dumps, command posts, etc.), or in the form of confirmation of pri soner-of-war reports, large scale movements or intentions of major enemy fortnations and the appearance of new battle units in a given sector. As an approximation of the use made of the data rece ived from Detachment "D", the CIRO PEARL, PEARL and THUMB included in the daily G-2 ISUM ... .The G-2 ISUM was sent by tele type, operational priority, to the following headquarters: SHAEF Main 6th Artny Group 21 Artny Group (Main) 8th Air Force First U S Artny Command Headquarte rs Third U S Artny Main Ninth U S ArnlY Main Fifteenth U S Anny Forward 12th Artny Group Main 12th Army Group Tac Headquarters First Allied Airborne ArnlY Com Z ASCZ 9th Air Force Advance d. It is of interest to note that an estimated 6,000 CIRO-PEARL messages were read during the seven-month period under di scussion. Of these, 3238 messages from twent y-fo ur diffe re nt formation s were publi shed in CIRO-PEARL/SSDD reports, which we re transm illed to G-2 , 12th Army Group. It is of further interest that on the average ei ther CIRO-PEARL, PEARL or THUMB mate rial , or all three, were included in one out of every three G-2 IS UMS . 5. SIGNAL RADIO INTELLIG ENCE AND SIGNAL SECURITY COM PANIES 12TH ARMY GROUP a. The mission of Detachment " 0 " in connection with the signal radio intelli gence units of 12th Anny Group was subdivided into three categories: ( I) Reporting o f Signal Inte lli gence of Immediate Value to Army and Corps G-2's (a) All tcchnical signal intelli gence of tactical value, particularly CIRO-PEARL, was transmilled da ily by the most expeditious means to the Army Signal Radio Intelligence Companies .... contains intelligence extracted from technicalmessages sent to the field units duri ng a typical month. This intelligence was used in two different ways: (l) It was reported to ArnlY and Corps G-2 's.

TACTICAL COMINT

195

(2) It was employed as a basi s for allocating intercept and direction-finding coverage priorities. (b) The most convenient means of transmission of technical signal intelligence to the field units was radio, teletype and courier. Two radio broadcasts were made daily, one in the morning and one in the early evening to establish coverage priorities and to assist in the preparation of technical reports for the respective G-2 Sections. Technical signal intelligence was also transmitted by the same means at any hour of the day or ni ght when the urgency of the information so warranted. (2) Consolidation of Technical Material of all Field Units (a) All technical data from the signal radio intelligence unit s of 12th Army Group were consolidated daily with those obtained by the two intercept units directly responsible to Detachme nt " D" and with data received from intercept units of the two nanking Army Groups (6th and 21). Items of immediate value were transmitted along with intelligence described in paragraph a preceding. In addition, all such data were published in greater detail in reports of the Intelligence Branch ... and dispatched to the field units. (b) The daily publications of the Branch not only aided the field units in the nornlal execution of their missions, but provided the only means of maintaining technical continuity on enemy formations during the frequent moves of the field units. This was particu larly true when Anny and Corps fronts were radically altered. rcsulting in an entirely new. and frequently unfamiliar, group of enemy units facing them. (3) Supervisory Control Over Operations of Signal Radio Intelli gence Field Units (a) Policies concerning operations emanating from higher headquarters, or those arising from within Detachment " D" from time to time, were distributed and enforced. Aid in complying with the policies was provided in the form of directives, special technical reports, and by conferences at the Detachment headquarters. Also frequent liai son visits to the field unit s were made by technical experts from the Intelligence Branch. 6. FLANKING ARMY GROUPS Continuous liaison was maintained with 21 and 6th ArnlY Group signal radio intelligence agencies both by pe rsonal visits and by radio teletype and telephone communications on a twent y- four hour bas is. All technical report s publi shed were exchanged with those two organizations.

SIGNALS INTELLIGENCE

196

From SRH 042. This accol/nt of signals intelligence operations in General

GeO/Re S. Patton's Third Army gives the reader a good feel for the "nl.lts and bolts" of World War /I collection activities ill the field. THIRD ARM Y RADIO INTELLIG ENCE HISTORY IN CA MPAIGN OF WESTERN EUROPE l. OPERATIONAL ORGANTZATION AND PROCEDURE OF RADIO INTE LLI GENCE AGENCIES. A. Organi zation of Signal Se rvice Companies. ( I) All signal service companies (corps signal radio intelligence companies) we re organi zed and equipped as prescribed by TIO & E I I -500 , I Jul y 1943, with an authorized strength of eight officers and one hundred twe nt y-one e nli sted men. Several weeks after activation, the 32S3 rd , 32S4th, 32SSth , and the 32S6th Signal Service Companies were moved to Wincham Hall , Eng land (K- I 9), in order that these companies might co llaborate on common problems of trai ning and supply. At a later date each signal service company was attached to a corps wi th whom it wo uld work in the field, and as the va rio us co rps we re committed in combat operations, the respecti ve corps compan ies wou ld accompa ny them remaining within a few miles of corps headq uarters in order that close liai son could be effected between the companies and the corps G-2. (a) After the usual training program in Burton Bradstock, England, the 32S3 rd Signal Se rvi ce Company left for Omaha Beach in France, arriving on 12 Jul y 1944. The company was attached to XV Corps and remained with Third Army until 2 I September 1944 when it moved unde r the command of Seventh Amly. (b) The 32S4th Signal Service Company was activated at Marbury Hall, Cheshire, England (K- I 9) on 19 April 1944. The operations of the compan y were begun immediately after activation; however,

it was not all an entire company

basis until operations against the e nemy on 2 July 1944. Prior to that time operations consisted of a general training program as we ll as individual team operations. A few intercept positions were establi shed in June at Lyme Regis, England, with the traffic analysis unit controlling the interce pt of nets and pre paring the intelli gence reports. On June 29 the company arrived in France with the primary miss ion of suppl ying tactical intelligence to VIU Corps. (c) Effecti ve 6 May 1944 the 32SSt h Signal Se rvice Company was activated at Marbury Hall, Cheshire, England. The unit has been with XII Corps since its activation except for brief periods in its earl y days when it was assigned to Anny so that activities of the company as well as the remainde r of the service companies cou ld be controlled and correlated by Army headquarte rs. After completion of the training program the compan y was concentrated at Lymington , England (U-

TACTICAL COM INT

197

7 1). and departed to Omaha Beach on 12 August 1944. Thus, the unit arrived in the combat zone ready to become operational on the ninty-eighth day after its activation in the Communications Zone. (d) Activated on 5 May 1944 at Wingham Hall , Cheshire, Northwich. England. the 3250th Signal Service Company was e ngaged in a signal intelligence training and field program for the nex t three months until 7 August when the company was committed at Courville-S ur- Huisine, France (T-08). Excepting for temporary attachment to III Corps during the Battle of the Bul ge, the company has always serviced XX Corps. (e) The 3259 th Si g nal Service Company came into ex iste nce o n 16 September 1944 under XIII Co rps. Training in Ringwood and Lymington. Hams, and Danford, Ke nt (Q-99) in England equipped the unit for its first ac tion unde r III Corps on the First U.S. Army front on 14 February 1945, and after the liquidation of the Ruhr Pocket, it was transferred with the III Corps to Third Army, whe re it remainded until the terminati on of the campaign. (f) At the cessation of hostilities on 8 May 1945, the following corps compan ies were with the Third Army: 3250th Signal Service Company under V Corps which was attached to Third Army for the last three weeks of the war; the 3255th and 32561h Signal Service Companies which served under Third Army during practically the entire campaign with XII and XX Corps res pectively: and the 3259th Signal Se rvi ce Com pany under III Corps. The 3254th Signal Service Company attached to VIII Corps al so spe nt muc h time under thi s command , although during the last wee ks of the war it was First Army. (2) I I 8th Signal Radio Intelli gence Company. The 1181h Signal Radio Intelligence Company was acti vated on 20 April 1942 at Fort Sam Houston, Texas based on T/O 11 /77 which was late r superseded by T / O & E II - SO~ , I July 1943, as amended b y Change No. I, dated 13 November 1943. After extensive training on maneuvers the company arrived in England in January 1944. At the direc tion of the Si gnal Offi cer, Headquart e rs European Theater of Operations United States Army, the inte lligence compan y, late r assigned to the Third Army. began intercepting German military traffic on 26 April at Dartford. Kent, between London and the Channel. During the nex t six weeks the practice traffic of the enemy indicated certain troop movements to and along the Atlantic wall, and also revea led radio characte ri stics late r used in identi fying and locating German units in the field. On I June, the compan y, then serving 21st Army Group, moved to a point on the English coast just across the Channel from the Cherbourg Peninsula where it covered e ne my radio activity during the invasion and part of the build-up period gaining slight inte lligence of operational value. On IS July the unit arrived in France and proceeded to the vicinity of the Third Artn y headquarte rs to whom it supplied intelligence through the e nd of the campaign. B. Radio Intelligence Procedures.

198

SIGNALS INTELLIG ENCE

(I) Ope rational Procedures. (a) Corps Level A ltho ugh the sig nal service compani es we re o rig ina ll y organi zed along identi cal lines, each, after being committed, adopted new methods of procedure, a nd changed the basic organization to fit spec ific o pe rational need s. T his report is designed not onl y to show the ori ginal composition o f these units, but also to point out the ir major differe nces. The companies included in thi s report are the 3253rd , 3254th, 3255th, and 3256th Signal Service Companies; although the 3255th and 3256th Companies were the only ones to be assigned to Thi rd Arm y throughout the entire campaign. the others were on separate occasions under thi s command .

Radio Inte rcept : The normal radio intercept secti on consisted of an inte rcept officer and from forty two to fort y six ope rators who we re di vided into tricks composed of a trick chief (control chief) and a num ber o f me n dete rmined by the traffic level of the c urre nt situati on. The o pe rato rs wo rked in HO l 7's and HO 27 's mounted on 2 1/2 ton trucks; the de ployment o f the trick chiefs varied with the company invo lved: The 3256th e mp loyed a trick chief fo r each truck but both the 3253rd and 3254th Companies used one man to contro l an entire inte rcept shi ft. In the 3255 th the use o f an inte r-commun icatio n system by whi ch the IA secti on could contac t the indi vidual ope rators entire ly elim inated the need fo r trick chiefs. T he average working day for the intercept sections was d ivided into three shift s with their times and set allocations as fo llows : 000 1-0800, eight sets; 0800- 1600, 15 sets; 1600-2400, fi fteen sets. The fun cti ons of the trick chief, whethe r he was in charge of the entire shift or of a single truck, were to control the inte rception of nets, check freque ncies. and in gene ral to act as an intemlediary betwee n the radio ope rators and the IA section. The process o f searching for nets was one which espec ially demanded contro l by a single source; the 3254th contro l was achieved by means of a home constructed monitoring board channeled to each ope rating pos iti on whe reas the 3256th simpl y subm itted a search list of known priorities to the indi vidual ope rato rs; these we re compiled by all radio intelligence sources by the IA Section on the second day of the German three day freque ncy cycle. This latte r method greatl y minim ized the ti me spe nt in reco rding nets of no immedi ate va lue, and thu s insured that the majorit y o f enemy units actively engaged on the front in question wo uld be covered. As a furth er aid, li sts of fixed call signs and pec uliarities o f procedure were also kept available at each set. In regard to the technical aspects o f radi o int elli gence wo rk. it was fo und that the supe r- pro rece ive r te nded to suppl ant the SC R 342 a nd 344, hav ing been adjudged as mo re adaptable to inte rcept requireme nt s. Eac h truck was usuall y equipped with one hi gh frequency set connected to an appropriate antenna. For both high and low frequency work, ante nnae o f varying characteri stics were ex pe rimented with by the individual companies: The hori zontal directional type proved to be the most popul ar with the 3255th , but the 3253rd Compan y on the othe r hand

TACTICAL COMINT

199

preferred the "L" shape antenna. Speak ing in temlS of general operating procedure it should be mentioned here that the interception of e nemy traffic by those companies e mployi ng the leap frog method of moving was rarely interrupted. When the tactical situation warranted. advance panies to division and forward headquarte rs were sent out and the proximity to the enemy lines proved reasonably profitable as demon strated by the extremely tactical low grade traffic intercepted by the 3255th and 3256th forward echelons. Since continuous wave interception was not the sole function of an intelligence agency, it was necessary to have on hand operators capable of copyi ng Gemlan voice traffic; these radio te lephony operators were part of the intercept personnel in the 3255 th Company, but in the 3256th they were furnished by the IA Section. The type of traffic obtained by this source was comparatively low in volume, and that which was inte rcepted ordinarily proved to be of local tactical value. Radio telephony personnel were also e mployed in the VHF (very high freque ncy) experiments when small teams were sent up close to the front equipped with high frequency intercept sets. The results obtained in these cases, as well as in the allempts to inte rcept German voice traffi c on the medium frequencie s from similarly advanced positions. were fairly good in so far as the mechanics of interception were concerned. but the intellige nce gained the refrom was not of far reaching val ue. Dircction Finding: Operations of the direction findin g teams in the signal service companies were for a long time hampered by technical diffic ulties involving equipment. The SCR 503 Receivers proved to be inadequate as long as the front remained fluid and even on a comparative static fro nt the direction finding sites had to be relatively close to the enemy in order to obtain satisfactory results. Later in the campaign the equipment was supplemented and modified by individual companies. and as a resull marked improveme nts were apparent. The final set used by the 3254th Company was an SCR 555 with an improved antenna loop and an SCR 244 as a receiver, the laller g iving a wider null but beller reception. The 3256th Company utilized Gennan eq uipment found at a signal depot in St. Menehould. France. but thi s too was effective onl y at short range. Generally, each company had two and some times three teams depending on the tactical situation ; one was always located at the company site and the other or others at some forward position. The conditions of the roads, the avai lability of water and ralions, the maintenance of radio and power eq uipment al the advanced locations. and the necessity of wire commun icat ion being in stalled for the most part by corps signal ballalions-all these factors had a direct bearing on the efliciency of the direction finding platoon. Cont rol of d irection finding stat ions was accomp li shed by various methods. The 3253rd Company placed a direction finding man directly in the radio trucks and acted as a liaison between the intercept and the direction finding sections. The

200

SIGNALS INTELLIGENCE

3254th Company on the other hand set aside one operator whose duty it was to pipe signals to the local direction finding station for re-piping to the advanced sites. In spite of all the difficulties encountered direction finding results from the standpoint of signal service companies were unifomlly good during the Bailie of the Bulge, the cleaning out of the Saar-Mose ll e Triangle, and the Bailie of the Rhine. Evidence derived from this source as a confirmation of the intelligence gained through other means consistently proved valuable. Traffic Analysis: The Traffic Analysis Section (IA Platoon) was commonly divided into two main groups, Traffic Analysis and Cryptanalysis Department s; originally the section had a complement of twelve men, but this number was later augmented by personnel from other platoons of the individual companies according to specific needs. Perhaps more appropriately referred to as the lntelligence and Coordination Team. the section had four principal functions: The analysis and processing of intercepted traffic and the identification of enemy radio nets; the cryptanalysis of enemy low grade codes and ciphers; the determining of Gemlan order of bailie and the maintenance of the necessary files on enemy units, personalities, code names, and other necessary data to achieve this end: lastly, the compiling of statistics on the technical aspects of the work to aid in the identification and analysis of the traffic. The Traffic Analysis Department usua ll y consisted of four traffic analysts whose duties were reading logs and searching the files in order to identify the current enemy units. There was a man designated to keep order of battle information including personality files , breakdown of enemy units, situation maps, and all such pertinent information derived from G-2 reports, captured documents, and other reliable sources. One of the analysts coordinated the information obtained from analysis and direction finding for subm ission to the traffic analysis officer for final approval: that which was of immediate tactical importance was relayed 10 corps G-2 and subsequent ly the finished product was dispatched to all other interested agencies,

Two tran sla tors and two c ryptanalyst s comprised the Cryptanalysis Department whose province was the breaking of all types of low grade traffic and the examination of all medium grade for possible compromises. When teletype communicat ion was available, raw traffic (low grade) was exchanged with the other intelligence units of Third Army. In some cases the men of this department were called on to interrogate prisoners thought to be of importance to radio intelligence work, to serve as radio telephony operators. and to process captured documents destined for higher headquarters. In this department as well as in traffic analysis it soon came about that each man, although trained for a specific position. became capable of handling a variety of duties. The plolling of direction finding bearings was usually done by some member of the Traffic Analysis Section, although it should be mentioned that much bCller results were obtained from assigning a certa in man to that duty. Most of the com-

TACTICAL COMINT

201

panies were restricted in their use of situation maps and overlays by the lack of space in the housing units. When the Traffic Analysis Section of the 3256th Company moved indoors, it was possible to exploit further the results of direction finding in connection with the order of bailIe picture. [n this company the direction finding man. in addition to his primary duties. maintained an overlay on a 1:25000 map showing the breakdown to ballalion level of all enemy unit s on the corps front; this map also included the direction finding bearings of the preceding day as well as any additional information of the same kind extracted from reports of other radio intelligence units . When it was poss ible to obtain a large number of bearings from various sources on any specific unit , a special overlay of that forn13tion 's area was made, thus affording a closer check on its movements. As a final comment all of the work done in evaluating intelligence gained from intercepted traffic was channeled to one principal destination, the daily activity report which served as a complete record of the day by day radio intelligence resu lt s. The form of the report was as fo ll ows: ([) Intelli gence Summary ; ( II ) Decodes and Translations; (1II) Technical Summary of Nets Heard ; ([V) Message Count and Set A ll ocation; (V) Direction Finding Bearings; (V I) Code Idents and Cipher Values. (b) ArnlY Level The Radio Intelligence Company at Army level naturally differed in several respects from the corps' agencies; approximately twice the size of the Signal Service Company, the Army unit had several additional functions, chief among which was the coordination of the work done at lower levels and the dissemination of information lhereto.

Radio Intercept: Aside from the disparity in numbers between the intercept section of the Army company and that of the corps (the fonner maintaining some twenty two men on each shift) there were major differences in operational procedures. Certain minor points nOl mentioned before were common to intercept work at all echelons. During times of heavy radio activity it was found best to have at least two illlercept operators on random search. who wou ld scan the dial ulllil an enemy net was heard. The frequency and call signs were then given to one of the other operators and the search was continued for other nets. When an enemy net went off the air, or when the operator. for various reasons. could no longer hear the transmissions, the spot checking practice was uscd. This meant that the operator would go on search, returning to the frequency every fifteen minutes. In order to ensure good copy of messages, important for the cryptanal ytic work. nelS were frequently double banked; since this meant the assigning of two or more operators 10 cover the same traffic. it was naturally more easi ly done at Army level. where more men were available. In any given case, the necessity of double banking was determined by the traffic analysis control man with the one exception.

202

SIGNALS INTELLIGENCE

that it was a standing rule to double bank all three letter traffic and plain text. In the case of voice transmi ss ion the handling of plain text in the Army company occasionally involved use of the dictaphone, the recording being sent to the Traffic Analysis Section. In certain urgent cases, the voice signal was piped by telephone directly to a translator in that section. When the Anny agenc y initially went into operations, the intercept and direction finding control man occupied different housing units and were linked by te lephone. It was later decided that the work of these two sections could be better coordinated if their control chiefs operated in the same unit and simultaneously received directions from the Traffic Analysis Section . Direction Finding: The size of this section and the existence in the Army company of a wire section to establish communications made it poss ible to operate as many as three outstations; later, radio control was used and proved more practicable than in the corps com panies. As a general rule, the length of the base line between the outside stations was forty miles with land line communications, and sixty miles with radio. As already mentioned, it was found best to have the direction finding control man work closely with the intercept trick chi ef, but aside from these points, the operational procedure involved in taking an y given bearing did not differ greatly from the methods used by the signal service companies. Traffic Analysis: The traffic analysis section of the Army company also differed in only a few respects from that at the lower level; of course, since this section was larger, being compri sed of twenty nine men and three officers, it was possible to extend the work of the various sub-sections to a finer point of detail and to act as a coordinator of the results obtained on the corps' front. As thi s latter function became more and more important , it became necessary to create a special section in the AmlY radio inte lligence company to act in liai son with corps, flanking armies, and Anny group; infomlation was thus routed more effi cientl y and expeditiously up and down the line. Security considerations, in many cases, made it imposs ible to send certain data do wn to corps level directl y and one of the fun ctions of the Army compan y was to see that the signal service units receiver whatever ex tract s and conclusions froIll

thi s information it was poss ible to disseminate. To further thi s end and to accommodate the large influx of the many different reports and documents, it was necessary at Anny level to create another sub-section within the traffic anal ysis pl aloon known as the incoming documents section ; the purpose of thi s miniature message center was 10 regulale the distribution of all incom ing and outgoing material and to keep the necessary eX lens ive fil es and records. In the matter of intercept and direction findin g control, the techniques were the same as those used by the signal service companies, except that the Army company was natu rall y ex pected to concern itself wi th all corps' fronts and even with some unils close to the Anny Zone of Advance in view of the possibility of sudden movement. The particular method of establi shing contro l generall y fo llowed one of

TACTICAL COM INT

203

those already outlined, that of distributing each day a li st of the desirable frequencies and call signs to the control chiefs of the other sections. Since the cryptanalytic section at Anny was larger than that at corps, it was soon found that in many cases rapid transmiss ion of traffic inte rcepted at the lower level to this section accelerated the eventual breaking of it. A teletype net involving Third Anny radio intelligence agencies prov ided the necessary means of communication and proved fairly practical. In thi s, as in alm ost every case, the g reater manpower in the Army traffic analysis section made possible much more detailed files and records of all kinds; as a result, intelligence could be inte rpreted on a wider basis. Whe n the data at AmlY level itself was insufficient, the Anny company was able to have recourse to Army gro up, o r. in certain instances. to Signal lntelligence Divi sion, European Theater of Operations United States Anny. This was especially important when a signal serv ice compan y suddenly found itself concemed with new Gennan units because of a re-a ll ocation of corps zo nes; in all probability, the Army company had already been intercepting traffic from these units, but if not, the above mentioned chan nel s for obtaining infonnation we re open. It can be readily seen, the refore, that the Anny Radio Intelligence Company was, in aLmost every respect, essentialIya larger scale signal service company with far more extensive functions because of its size and position in the organization of a Field Anny. (2) Communication Set-Up of Radio Intelligence Agencies . ... Signal intelligence agencies empl oyed most of the known means of communication . Each compan y Mobile Radio Team operated a sub-station of a net with the AmlY Radio Inte lligence Company as control. The AmlY company itse lf had a lateral radi o link with its corresponding unit in flankin g armies, and of course was part of the Twelfth Anny Group Radio Intelligence Net. In the signal service units radio contact with direction findin g o utstations did not prove to be quite as satisfactory (in compari son with wire lines) as at Anny level. For communication within the companies themselves telephone naturall y was used, but the re we re some variations in the detail s of its use; the 3255 th Company and the I I 8th Signal Radio Inte lligence Company had their system s so arranged that each individual operator and trick chief could communicate separately with the traffic analysis section , but in the 3256th Company direct communication ex isted only betwee n the trick chiefs in each truck and the IA inter-communication ope rator. A teletype net was operated for a long time involving the signal serv ice companies, the Army compan y, and Third Army Signal Intelligence Service. The chief purpose here was the exchange of raw traffic , and, on several occasions, thi s means

was employed to send traffic back to the compan y prope r from advance teams. When the teletype lines were in good order, thi s was by far the quickest, easiest, and most secure manner of exchang ing info nnation. During certai n periods the re was a teletype line betwee n Army group and the I I 8th Signal Radio Intelligence

204

SIGNALS INTELLIGENCE

Company over which raw traffic and items of technical infonnation were transmitted. Telephone was naturally widel y used. Both amlY and corps agencies communicated with their respective G-2 scctions by thi s means and in addition the fonner always had a direc t line to the Signal Intelligence Service Office at Army headquarters. This latte r circuit was supple mented by a direct tele type line to provide printed confirmation of any information relayed to G-2 by te le phone. Telephone communications further existed between the Arm y company and Army gro up , flanking armies, and the signal service companies in Third AmlY. Whenever a telephone or teletype line passed through one or more switchboards, a jargon code was used , except in the te le type transmission of raw traffic. C. Captured Documents. From the beginning of the campaign in Nonnandy the importance of captured documents to increase the value of radio intelli gence infonnation was ascertained. During the first few weeks after the initial landing. linle was done in thi s direction because the static line prevented the capture of documents in any quantity. After the St. Lo breakthrough advantage was take n of an increased numbe r of captured documents which necessitated a system of di stribution , so that thi s information could be expcditiously disseminated. Corps' G-2 first analyzed any documents in search o f fact s peculiar to its own work, then passed all signal docume nts on to the corps ' signal service companies for analysis. Any infonnation of value was immediately extrac ted and the original document was sent to Anny Signal Intelligence Service where it was photographically reproduced and distributed to service companies under Third Army Command, First, Seventh, and Ninth American Armies and Twelfth Army Group. Code names, code and cipher keys, frequency and call s ign allocations, maps with encoded terrain features and place names, map coordinate codes and grid systems, and the German me thods of signal intelligence operations constituted information of partic ular impOltance. Over three hundred signal intelligence doc ume nts requiring some eight hundred pages were re produced and disseminated during the course of the campaign. To supplement this infomlation signal intelligence notes were published and distributed when document s received were particularly current or when only extracts from lengthy documents were of value. Enemy units were often identified through code names continued in document s, and encoded traffic was frequently easily read due to the capture and dissemination of documents, diagrams, and German signal intelligence

publications. During August in the early days of the campaign, intelligence companies were able to follow the movement of the 9 Panzer Division from non-indicator cipher keys captured from thi s unit. Numerous examples could be c ited to justify the work required in maintaining a small Documents Evaluation, Reprodu ction and

Dissemination Section . D. Prisoner of War Interrogations. In order to obtain all available radio intellige nce infonnation PW interrogations were made frequently. During the campaign the various PW cages were

TACTICAL COMINT

205

visited at leas t once a week under ordin ary circum stances . Arran gement s we re

made wi th PW autho rities that Signal Intelligence Service should be notified when routine screening indicated that a PW might possess informati on of signal intelli gence value. In most instances the information obtained from PW 's had alread y been secured through other sources. however, an y additional facts al ways served as a confirmation. Frequentl y PW 's were interrogated soon enough aft er capture to obtain detailed informati on on effecti ve frequencies and call signs which otherwise was a difficult mailer in view of the fact that they, excepting in most cases of reconnaissance traffic, were customaril y changed every third day in the Gernlan army. Although valuable radio intelligence infornlation was obtained from PW interrogation, its principal virtue lay in the securit y angle. German signal intelli gence men, especiall y those fTom corps headquarters and higher, could always give information as to which of our code or cipher systems were be ing broken, what diffi culties arose in the process, how long a system could be depended on for security, and what disposition was made of the inte lligence gained . ...

*** From SRH 227. This is parI of Ih e official unil hislOIY of a signal radio inlelligence company assigned 10 Ih e S oulh weS I Pacif ic A rea , General DOllglas MacAnhur's Ih ealer of operalions. II should be nOled Ihal IInlike Ih e silllalion in Ellrope, where laclical COMINT unils concenlraled Ih eir efforts on low-level lraffle, Ih e 126 1h Signal Radio Inlelligence Company repon ed direclly 10 Ce ri/ral Bureau, a high-level Ihealer processing cenler. Th e informalion in Ih e fi rsl senlence is aclt/ally incorrecl: Ihe unil's predecessOl; Ih e l SI Radio Inlelligence Company, was aClivaled ill 1938 and was 1101 a provisional unit. OPERATIONAL HISTORY I26TH SIGNAL RADIO INTELLIGENCE COMPA NY February 1944 The I st Si gnal Radio Intelligence Compan y was acti vated at Fort Monmouth , New Jersey, in February 194 1. having prev iously been known as the I st Prov isional Radio Intelligence Company. The operational set-up of the new company inc luded three operating platoons, each of whi ch was to be capable of operating independentl y on a combined intercept, directi on findin g, wire-laying mission. Operational train ing was begun with regular classes in International code, d irection find ing, plotting, and wire-laying. 206-C Directi on Finders were set up under simul atcd

206

SIGNALS INTELLIGENCE

field conditions and bearings were taken on local stations, to familiarize the operating teams with the methods of operation. In March 1941 the company moved to Fort Meade, Maryland, where training was continued. In June of the same year. the unit designation was changed to the 121 st Signal Radio Intelligence Company. The company participated in the North and South Carolina maneuvers during the period extending from September through the 6th December 1941. Intercept positions were set up in a 2-1/2 ton truck which also included a typewriter at each of the eight intercept positions. Back to back with the intercept truck was another 2-1/2 ton truck containing the plotting equipment. Goniometric wire lines were laid out to the four 206-C direction finders, which were the only OF's in the company with the exception of the 504 equipment items, which were not used. The direction finders were generally spread over a base line of approximately 20 miles. Thus, at this stage, with limited equipment, the company was not using the three-operating-platoon set-up. The company worked at one time with the Red army, at another with the Blue army, intercepting "enemy" traffic, and locating "enemy" transmitters, for GHQ, in each case. The company also worked with the referees, monitoring all radio traffic for violations of signal security. Following maneuvers, the company returned to Fort Meade, Maryland . Shortly thereafter it moved again to Ft Bragg, North Carolina where it participated in dayto-day maneuvers with the field artillery, locating "enemy" transmitters, which were than " neutralized" by the artillery. Following the company's return to Ft Meade, on the 25th of February 1942, one operating platoon of the 121 st Signal Radio Intelligence Company was designated as the First Operating Platoon. 1st Lt Menear was placed in command of the platoon over 45 enlisted men. The platoon left Ft Meade for Ft Dix, New Jersey, LO await movement orders for an overseas destination. Its equipment included intercepL receivers, the four 206-C direction finders , plotting equipment, wire and wirelaying equipment, and field range equipment for independenL operaLions. On the I SL of March 1942 the FirsL OperaLing PlaLoon sailed from the Brooklyn Port of Embarkation on the USATS URUGUAY. On board ship the plaLoon radio operators were placed in charge of the radio room and the signal bridge, and operated the ship's radio, blinkers, and signal flags. The ship arrived at Melbourne, Australia, on the 10Lh of April 1942. AL Camp Pell, Melbourne, training in intercepL and direcLion finding operations was continued until the 6th of June. Meanwhile, on the 19Lh of May 1942, the platoon was redesignated, and became the First Operating Platoon, 126th Signal Radio Intelligence Company. On the 6th of June the plaLoon left Camp Pell and proceeded by motor convoy to Mt. Massadon, Victoria, the contemplated site for a pemlanent camp. The platoon set itself up for independent operation, performing its own house-keeping and cooking chores. InLercepL faciliLies, consisLing of three posiLions, were operated in

TACTICAL COMINT

207

the van of a 1-1/2 ton truck. One direction finder, 206-C, was placed in operation atop Mt. Massadon , about 3/4 miles from the camp. On I July 1942, while still at this location, TISgt Messer of the 823rd Signal Service Company was attached to the platoon for temporary duty to instruct the platoon's intercept operators in ROMAJI and Japanese KANA telegraph code. On 10 July 1942, the platoon moved again by motor convoy to Townsv ille, Queensland, arriving there on 2 August 1942. Lt Menear was reported to Capt Brown, GHQ radio intelligence officer, who in turn assigned two GHQ officers, Lt Phelan and WIO Card , to establish a Japanese code school for the instruction of the platoon 's intercept operators. In addition, the platoon was assigned the mission of monitoring Allied Air Reconnaissance frequencies copying Allied clear and coded messages during the attacks on Guadalcanal and reporting by TTY to GHQ. The platoon 's 206-C direction finders were set up to work with the No I Wireless Unit, RAAF, which was operating in Townsville. The OF's were located respectively at Townsville , Cairns, Charters Towers, and Cloncurry, and communicated with Townsville by wire ... .Their mission was to work with the RAAF unit in securing bearings on enemy aircraft. However, the work was not wholly successful because of the long lines of communication and because the OF's were not suited to long range work, nor to the rapid action required to obtain bearings on short aircraft transmissions. In addition, the OF operators were not trained to read Japanese code, and therefore could not distinguish readily between the various transmissions they were to pick up. On 15 November, the platoon travelled again , thi s time by rail to Brisbane, Queensland. MISgt Nurss, working under Capt Brown and GHQ, had preceded the platoon to arrange for the establishment of a radio station and antenna equipment. The platoon arrived in Brisbane on 18 November 1942 and immediately began work on a new operational site at Northgate. With the completion of a radio "shack " constructed by Australian workmen, and an antenna installation, using 6 rhombic antennas. work was begun in intercept, covering Japanese Army and Diplomatic radio traffic. All traffic was turned over to Central Bureau , GHQ, for analysis. Capt Brown , and his officers, Lt Phelan and WIO Card directed the operations of the platoon from this point on, using experience gained in and since the invasion of the Philippines to further the instruction and efforts of the platoon's intercept operators. It was necessary to train more men as intercept operators. Consequently, all men, regardless of their previous assignments in the platoon were given training, and of these the best became intercept operators. No direction finding work was done during this period. The rainy season which followed shortly thereafter caused the si te of the radio station to become inundated. On 10 January 1943 the platoon moved to Stafford. Brisbane, where it established an independent camp and a temporary radio station, the latter housed in two tents, and using simple antenna equipment. Work on a new GHQ radio station was begun at once, a short distance from campsi te. Australian

SIGNALS INTELLIGENCE

208

workmen constructed the building. Grounded V-antennas were installed to cover directionally the arc from northwest through northeast of Brisbane. Fifteen antennas were set up for this purpose, arranged in groups of three, each group covering the same direction with spacing for diversity use if necessary. By the beg inning of March the new radio station was in operation, continuing the work beg un at Northgate. The amount of enemy traffic being copies by the First Operating Platoon, 126 Signal Radio Intelligence Company, was almost being doubled each month. In April 1943, the First Operating Platoon , all its personnel except Lt Menear, and all its equipment, were combined with the 126th Signal Radio Intelligence Company which had arrived in Brisbane on the 25th of March 1943. The 126th Signal Radio Intelligence Company had its origin at Camp Crowder, Missouri , where it was activated on the 14th of August 1942. Five officers, all second lieutenants, were present for duty with the company on the date of its activation. By virtue of seniority, Lt Joe L. Thurston assumed command . . . .The strength of the organization on the date of its activation, was 5 officers, 0 warrant officers, and 85 enlisted men . The Table of Organization of the company included a group known only as the First Operating Platoon , which had already been activated. The whereabouts of the First Operating Platoon was not known or not revealed. Howeve r, its strength of I officer and 45 enlisted men was to be inc luded as part of the stre ngth of the 126th Signal Radio Intelli gence Company, and it was believed that the two units would eventually be joined . The new company was placed under the Signal Corps Unit Training Center at Camp Crowder, for administration and preparation for active duty overseas. It was the function of the Unit Training Center to coordinate administration, training, and securing of personnel and equipment for newly activated SOS units at Camp Crowder. On the 19th August 1942, followin g movement of the company to it s own area at Camp Crowder, operational training began. Full advantage was take n of specialists' school s available at the MidWeste rn Signal Corps School and the MidWeste rn Signal Corps Replacement Training Center at Camp Crowder. Chauffeurs and fi eld linemen we re sent to schoo l for further training. All radio operators we re enrolled for advanced training in Inte rnati onal code, Joint Anny and Navy Procedures, and net operations.

Among the radio operators assigned to the company were twent y spec iali sts, wi th kno wledge of the Japanese Romaji charac te rs and thei r code eq ui va le nts. These men had received the ir special training at Fort Monmouth , New Jersey, and had been assigned from there to join the company on the 19th August 1942. At that time the re was no course of training at Camp Crowder suitable for specialists of this type. For the time being, therefore, they were entered with the other operato rs in the Inlernati onal code classes.

***

TACTICAL COMlNT

209

Four SCR-255 Direction Finders had been rece ived by the company. These were set up in the field near camp and connected by goniometric wire lines for several days operation and practice. Tentative operating teams for the OF's had been assigned for thi s work to learn as much as possbile about the equipment, and to use that knowledge in teaching others . All officers familiarized themselves with the equ ipment. Lt Dibos took charge of assembly and disassembl y, and instruction in OF operation. The direction finders were then dismantled and prepared for crating. Additional supplies and operational equipment were received by the company during the remainder of the month of September. Lt Kuch was tentatively placed in charge of intercept training. A schedule was arranged whereby the twenty Japanese specialist operators were to work several hours each day on company receivers installed in one of the barracks, picking up and copying Japanese radio code transmissions. This procedure was not too sati sfactory due to weakness and unreliability of the signals. Therefore, training was later amplified to include copying of Japanese code from disc and tape recordings, made for the purpose by one of the operators. During all thi s period, activities of the company were centered around crating and preparing for overseas movement. In early October 1942 , news was received of a short delay, due to supply shortages and changes in War Department plans. The company, however, remained alerted . One direction finder SCR-255 was uncrated and assembled near the camp. OF crews were assigned to work under the instructi on of non-commi ss ioned officers who had previous training, in assembling and di sassembling the direction finder, as well as in operating. SCR-284 transmitte rs were installed, both as fixed and as mobile stations, and operators were trained to tune and operate these sets; 342-type receivers were used to intercept the transmission of the SCR-284 's for the practice of copy ing and guard ing friend ly transmiss ions. Lt Skinner was transferred from the company on the 21 st of October 1942 due to a phys ical ailment; and Lt Hadd replaced him as Supply and Motor Transport officer. By oral assignment , Lt Dibos became Operations Officer, Lt Wil son was made Liai son and special services officer, Lt Swom ley became Wire and Personne l Officer, and Lt Kuch was assig ned as Intercept Officer. During November 1942 word was received of a further delay in overseas movement. Training cont inued through the use of Signal Corps specialists' schoo ls ava ilable on the Post. New intercept operators were bei ng assigned to the company from time to time, and these were entered in new code and traffic classes, which were espec ia lly arranged for throu gh the rad io sc hoo ls of the Replacement Training Center. Men who were not operators, but who were be li eved to ha ve potentialities as such, were entered in beginners ' classes. Many of the men assig ned to the company as speciali sts in certain types of work, were found to be inadequately trained . Consequently, eight men were selected to attend night c lasses in teletype operation; five men who had been assigned to the company as radio

SIGNALS INTELLIGENCE

210

repair men were sent to the repair section of the Signal Corps Rep lacement Training Center, to learn and work on maintenance of Signal Corps receivers and transmitters; and the wire platoon was assigned to practice pole-climbing, use of the compass and maps in nearby fields and woods, and use of hand and engine-driven reels.

*** During early December crating was again speeded in anticipation of overseas movement. Training was continued spasmodically, due to the need for men to assist in packing and crating. The Unit Training Center began a new training project, in which it created an "army" headquarters, and two "corps" headquarters, with emphasis on message center work. All available intercept, telegraph, teletype, and switchboard operators of the 126th Signal Radio Inte lligence Company were assigned to operate with other groups in the three headquarters' message centers. Radio operators handled radio traffic between the headq uarters, using Internat iona l code. Several company receivers. operated by company personnel, were set up and used as army and corps intercept units, guarding the radio nets and copying all traffic. After the first week, however, orders were received to ship all company equ ipment to arrive at the San Francisco Port of Embarkation by I January 1943. On the 23 rd of December, the eq uipment was shipped in three box cars from Camp Crowder. At this time, twenty enlisted men were enrolled in a six weeks cou rse of training which included operation of the 206-D and SCR-255 direction finders , elementary map reading, e lementary maintenance of radio equipment, e lemen ta ry trigonometry, elementary wave-propogation, and the technique of recording on cy linders and tapes. The men selected for thi s course were NCO's of the operati ons section, nco's in charge of directi on finding teams, as well as direction finder operators who showed potentialities fo r poss ible leadership of teams. During January 1943, the company remained alerted, although movement was again delayed. Toward the end of the month the company was given a final inspection by a staff of officers and me n from the Inspector General 's De partm ent. During the last week in January, orders were received for personnel of the company to move to the port of embarkation. All men in school were withdrawn. Lts Wil son and Hadd were sent immediately to the port to act as liaison officers for personnel and supplies respectively, and to prepare for arrival of the main body of the company.. .. On the 3rd of February 1943, the company departed by railroad for the San Francisco Port of Embarkation, arriving there on the 6th . The compan y was assigned for a period of staging and physical " harden ing" to quarters at Camp Stoneman. Then, on the 25th of February, with a strength of 6 officers, I warrant officer, and 203 enli sted men, the compan y boarded the USATS "Pres ident

211

TACTICAL COMTNT

Johnson" for shipment overseas. The company disembarked at Brisbane on the 23rd of March 1943, ,md marched to Camp Doomben, where casual quarters were assigned. During the third day at Camp Doomben, the company was visited by GHQ officer, Capt Howard M Brown, and by 1st Lt Menear. The latter was introduced as the commanding ofFicer of the First Operating Platoon for which allowances had been made on the Table of Organization, and the location of which had until then been unknown to the company....

From SRH 049. This document was produced by the 12th Army Group's Signal Security Division "D". The excerpt shows the results of one day's intercept of low-level German communications by Army tactical COMINT ullits. 011 November 11, 1944, intelligence was produced on three German divisions: the 151h Panzergrenadier Division ; 17th SS Pan:er Division; and 116th Panzer Division. Intelligence came from exploitation o.f German messages enciphered in medium-level cryptosystems (c/RO PEARL); low-level systems (PEARL); and radio direction-finding (THUMB). As can be seen from the text, CIRO PEARL messages were illtercepted on November 6, but could not be decrypted and translated IIl11il November II. The "cut" referred to in the jillal paragraph describes simultaneous bearings on an enemy radio station obtained by two separate radio directionjinders.

TECHNICAL SIGNAL INTELLIGENCE TRANSMITTED DIRECTLY TO G-2, 12th ARMY GROUP, ETO FROM 14 AUGUST 1944 TO MAY 1945

*** II NOY 44 ClRO PEARL 15PG DIY 1159 6/11 To \la Query whether tank specialists, who were recently sent here from 5 (PZ) Co are to be retrained to become Panzer Grenadiers. 15 PG DIY 6/11 1219 T02Co Send Rasterschluessel for course of instruction at FEB to FEB via registry. HORN

SIGNALS INTELLIGENCE

212

IS PG DIY 1417 6/11 From actual strength status of I Nov 44 (with exception of the sick and wounded within the past 8 weeks, as well as designated Fahnenjunker (OCS candidates), 10 percent NCO 's and OR 's of the age groups 1906 and younger fit for general service are to be sent enroute to FELD ERS BN by 7 Nov 44 at 1200 hrs in accordance with OKW order. Numbers of NCO 's and OR's released to Feb is to be reported to lib by 7 Nov at 2000 hrs. Written orders enroute. !la IS PG DIY 1531 6/11 SECRET - Report is to be made on Nov 10 as to whether there are experiences concerning rocket bombs having been shot down by pursuit planes. PEARL 17 SS PZ DIY Msgs today were mainly concerned with the difficulties of an art grp NAGLER (17 SS PZ ART RGT) which was in posi tion the morning of 9 Nov in ALEMONT (U 9040) . At 101410 the art group urgently informed its control station that Allied forces were advancing on BEAUX (U 9645), its present position, and requested immediate orders. At 101455 the group was told in answer to an inquiry that there were no munitions. At 101455 the group was still in BEAUX impatiently awaiting orders. At 101506 it was told that SCHMIDT (poss CO 10/38 SS PGR) was cut off and was requested to hold its present position. Shortly afterwards, however, the group indicated it was moving NW near PONTOY (U 9447). 116 PZ DIY 10/ 11 Rgt Bn of Art Rgt 146 still at RATH (F 10753 1) 116 PZ DIY 10/ 11 0345 Barrel s already dismounted. Ammo packed. Phone comms di smantled. Prime movers are in the btry position. Equipment loaded. Query - are we to prepare to fire again and is the OP to be reestablished. SCHWEIGER 116 PZ DIY 10/ 11 ENGELS (PZ AR 146) arrived with 4400 liters. THUMB 116PZDIY DfF on supply elts of 116 PZ DIY confirnl northward movement with cut on Art net at K 893742.

****

CHAPTER X Reshaping the Tools By 1945, there were growing pressures to rationalize the Anny's signals intelligence structure. A major step towards achieving this had already been taken in December 1944, when, as we have seen, the Military Intelligence Service assumed operational (but not administrative) control of the Signal Security Agency. Another initiative came about as the result of a belated Navy proposal to finally create an Army-Navy Communications Intelligence Board to coordinate the activities of the two service signals intelligence arms . Major General Clayton Bissell, the Army Assistant Chief of Staff for Intelligence, initially demurred, on grounds that the Army 's structure was too decentralized to allow for meaningful interservice collaboration. However, he then used the Navy approach as a wedge to establish centralized Army control over all it s COMINT assets. His arguments in favor of this are set forth in the excerpt from SRH 169, Centralized Control of U.S. Anny Signal Intelligence Activities. With the war in Japan coming to a close, all the theater commanders in the Pacific-even Douglas MacArthur-accepted implementation of the proposition once the fighting was over. On 15 September 1945, just thirteen days after the war in the Pacific had ended, the Army created the Army Security Agency, which would control all Army signals intell igence elements through a vertical command structure for the next 32 years.

214

SIGNALS INTELLIGENCE

From SRH 169. The head af the War Department's Military Intellig ence Division argues that all signals intelligence assets should be under his control. His appeal was ultimately successful. CENTRALIZED CONTROL OF U.S. ARMY SIGNAL INTELLIGENCE ACTlVITfES

*** I June 1945 MEMORANDUM FOR THE CHIEF OF STAFF SUBJECT: Centralized Control of Signal Intelligence Activities for War Against the Japanese I. On 8 May the War Department requested comments of the U.S. Army Theater Commanders in the Pacific and Far East on a proposal to place under the control of a single War Department agency all Army units and personnel engaged in signal intelligence activities for the war against the Japanese . ... 2. General Wedemeyer has advised of his "wholehearted approval" of this proposal. ... General Sultan advises that he "concurs in the need for a central agency to coordinate" all U.S. signal intelligence activities but believes this can be accomplished "without removing units and personnel from theater commanders' control" . . . .General MacArthur advises that he does not concur in the proposal because it would place such activities under the "absentee control" of an agency "many thousands of miles away"; would interfere with the "close association" which his headquarters has had with such activities; would "disrupt an Allied agency which is now functioning most effectively" in hi s theater and would prevent signal intelligence results from being made available to the theater commander with the required speed .... 3. Attached hereto is a memorandum which explains the facts and circumstances pertinent to the proposed centralized War Department control. It is believed that such control is essential: a. To coordinate U.S . Army signal intelligence activities in the various theaters and in the United States, eliminate duplication of effort and in sure full exploitation of signal intelligence sources. b. To coordinate U.S. Army signal intelligence activities with those of the U.S. Navy and the British. c. To utilize most economically the limited Japanese linguistic and other specialized personnel available for this work. d. To provide full technical support for all advanced signal intelligence units.

RESHAPING THE TOOLS

215

e. To ensure maximum distribution and utilization of signal intelligence technical data and results regardless of place of production. 4. The objections raised by General MacArthur are not considered valid. The world-wide nature of enemy radio il1lerception and the total absence of any relationship between Theater boundaries and the location of interception or solution of Japanese messages, makes it essential that the principal cryptana lytic center be at a geographic point where the machinery, facilities and skilled research personnel are located if the signal intelligence system is to produce the maximum intelligence of which it is capable. Control should be centrali zed at this same point which is the on ly location that has or can have access to all of the enemy radio traffic without prohibitive duplication of personnel and facilit ies. The principal cryptanalytic center is near Washington and must remain here because it would be totally impracticable to move the machinery and facilities to another site and secure effective results during the probable period of the Japanese war. Adequate communications exist to ensure not only fullest exploitation , but to guarantee almost instantaneous delivery of results to any Theater Headquarters. RECOMMENDATIONS 5. It is recommended that: a. The proposed centralized contro l of the signal intelligence system be approved. b. The attached draft of cable to thealer commanders concerned be dispatched . CLAYTON BISSELL Major General, GSC Assistant Chief of Staff, G-2

*** I. Circumstances leading 10 propasal I. The present recommendation for centrali zing control of all U.S. Arnly signal intelligence activities for the war against Japan is a direct outgrowth of a proposal made by Admiral King on 14 February 1945 to establish an Army-Navy Communication Intelligence Board charged with coordinating the plans and operations of the communications intelligence organizations of the Army and Navy.... 2. All signal intelligence activit ies of the Navy, including those conducted within theaters of operations, are under the direct operational control of the Navy Department in Washington ... .In the ArnlY, on the other hand, responsibility for signal intelligence activities is divided among the War Department and theaters of operations in the same manner as responsibility for the conduct of strictly combat operations against the enemy. Although this basic difference in the signal intelli-

216

SIGNALS INTELLIGENCE

gence organizations of the two Services made it unlikely that the proposed ArmyNavy Board could function as effectively on the Army side as it could 011 the Navy side . .. , that fact was not considered an adequate ground for rejecting Admiral King's proposa\. ... The proposal was therefore accepted by the Chief of Staff.. .. 3. However, Admiral King's proposal gave impetus to consideration by the Deputy Chief of Staff, the Military Intelligence Division and the Operations Division of the need for centralizing control of the Army's signal intelligence activities for the war against the Japanese. This consideration led to the conclusion that the basic principles of command applicable to combat operations are entirely inapplicable to the highly technical and specialized activities involved in deriving intelligence from Japanese military radio communications, and that all such activities should be placed under the control of a single War Department agency .... II. Proposal 10 theaters and replies thereto 4. Accordingly, after receipt of an independent suggestion from General Sultan as to the need for centralizing control of signal intelligence activities ... and after obtaining the views of the Chief Signal Officer..., the proposal now under considerarion was presented for comment to the theater commanders concerned . .. 5. The comments received from the theater commanders have raised no COIl-

siderations not previously taken into account by the War Department. General Wedemeyer's approval of the proposal is based upon considerations similar to those which constituted the basis for the proposa\. ... General Sultan 's reply approves the basic idea of centralized control and suggests only a different procedure for putting it into effect. ... General MacArthur's reply... raises objections of the same nature as those which had previously been considered and rejected as invalid .... Those objections are discussed below. "' Basic reasons/or proposal 6. The basic reason for the present proposal is that Japanese radio communications, their availability for interception , the cryptographic systems used. and the usefulness of the results derived therefrom, all have no relation to the considerations which are determinative in assigning command responsibility for combat operations against enemy forces.

Availability o/traffic/or interception. 7. There is no relation between the availability of Japanese military traffic for interception and the geographic organization of our ArnlY for combat operations. Traffic of great intelligence value to one theater may be intercepted at a point thousands of miles from the theater. For example, more of the Japanese military traffic received by Signal Security Agency is now intercepted in California than at any other place . The most advantageous locations for intercepting Japanese military traffic originating in Manchuria, Korea and North China are in the vicinity of Manila, Guam and Adak; the intelligence derived from this traffic may be of operational interest to any of the Far Eastern theaters, the Twentieth Air Force and the War Depallment.

RESHAPING THE TOOLS

217

Cryptanalysis 8. The main Japanese cryptographic systems are used throughout the Japanese Army. Itext withheld] High level Japanese systems can be solved with the use of large numbers of personnel and elaborate tabulating machinery. The degree of success depends directly upon the volume of traffic obtained; no one theater intercepts sufficient volume for continuous high level cryptanalaysis. Effective exploitation requires the pooling at one place of all available traffic from all available U.S. and Allied intercept facilities, regardless of their location. This place might be at almost any location in the world where adequate personnel and machinery could be assembled and 10 which adequate communication facilities exist for forwarding traffic. At present, the principal such cryptana lyti c agency is Arlington Hall Station, Arlington, Virginia, operated by Signal Security Agency under the operational command and control of G-2. 9. Forward exploitation of traffic at or near the point of intercept is desirable only where (I) the current period of the cryptanalytic system has been solved at the central cryptanalytic center having all traffic available to it, and (2) communication facilities are not adequate to permit particular messages in the solved system to be transmilled as quickly to the central agency as they can be to a local exploitati on center. Because of the almost instantaneous speed of electrical transmission, mere distance in itself has no bearing upon the adequacy of communications; they may be and often are beller to a point thousands of miles from the point of intercept than they are to a point much nearer. 10. The above basic considerations make it imperative that this work be under centra l direction and contro l. Any division of responsibility for this work among the War Department and theaters can have lillie if any relation to the practical problems involved, and necessarily results in duplication and inefficiency, without compensating advantages. Use/uilless 0/ results ohtained. II. If it were allempted 10 divide interception and cryptanalysis of Japanese military traffic according to U.S. Army theater boundaries. the results obtained within each theater would have little if any relation 10 the combat operations conducted in that theater. From the cryptanalytic point of view each theater commander would be forced to exploit all traffic which he could gel. regardless of its place of intercept or its value 10 his own operations , in order 10 read any traffic of primary concern to him.This would necessitate each theater commander duplicating the entire effort of other theaters and the War Department , in order to produce results most of wh ich would be of no greater value to his theater than to other theaters. For example, the Alaskan Defense Command and the China Theater might produce results which wou ld be of greatest value to the Southwest Pacific if that theater were to undertake operations on the coast of China or in the main Japanese islands. On the other hand, the Southwest Pacific might produce resu lts most useful to U.S . Commanders in China for directing operations of Chinese forces from

218

SIGNALS INTELLIGENCE

the interior of China. Either theater might produce results of primary interest to the Twentieth Air Force or to the War Department. "Low leve/" signal il1lelligence. 12. The foregoing comments are applicable chiefly to "high-level" signal intelligence activities, i.e., intercepting and reading Japanese messages transmitted in codes and ciphers having a high degree of security classification. Similar considerations are applicable to the important field of " low-level" signal intelligence, i.e., reading communications transmitted in low-grade codes and ciphers, or obtaining intelligence from the observation of radio traffic without reading the texts of the messages (traffic analysis). 13. There are very few Japanese cryptographic systems which are " low level " and those low level systems which it has been found possi ble to exploit are used in widely separated areas. For example, the most prOductive low level Japanese system now being exploited is an air-ground system used by the Japanese throughout the Pacific and Far East. Most effective exploitation of that and similar low level systems is impossible without immediate and full exchange of current technical data and results among the various signal intelligence units throughout the Pacific and Far East. 14. An additional important factor peculiarly applicable to the Japanese problem is the need for close integration of the resu lt s of this low level work with the results obtained from the high leve l work done primarily by Signal Security Agency. Only by a close tie-in of result s from all sources can complete and reliable intelligence be produced. 15. For these reasons, a central agency is required in order to afford each unit working on this low level material the full est technical support and to integrate the results of its work with the results obtained from other sources. IV Objections raised by SWPA 16. The principal objection rai sed by SWPA is to "absentee control" by an agency "many thousands of miles away," with "such delays as to seriously reduce the value of the derived intelligence and in many cases render it completely useless." These objections ignore the basic nature of signal intelligence activities and refl ect a desire for theater complete self-sufficiency which does not and cannot ever exist due to such limiting factors as critical maChinery, trained cryptanalysts, linguists and skilled research personnel. 17. The objection that the War Department agency is "many thousands of miles away" di sregards the speed of modern electrical communications. This is strikingly illustrated by the fact that the Director of the Briti sh Government Code and Cipher School has recently advised the War Department that the British are greatly expanding their intercept facilities in India and Burma, but that they desire all of the traffic intercepted in those areas (which are as far from Washington as SWPA) to be exploited cryptanalytically only in London and Washington (w ith basic research on certain systems done in London and primary reading of current

RESHAPING THE TOOLS

219

messages done in Washington). The Director recommends against exploitation by the Briti sh personnel of the joint agency in SWPA. [text withheld\ It recognizes that the important question is not at what location the exploitation is done so long as it is all done at one place. 18. The degree to which all exploitation should be concentrated at a single center depends not on the location of that center but upon the availability of communications between that agency and (I) all places of intercept and (2) all headquarters needing the derived intelligence. Experience has demonstrated that most Japanese military traffic now intercepted in the Pacific and Far East can be forwarded to Arlington Hall Station, decoded, translated, evaluated and disseminated to operational headquarters more quickly than the same processes can be performed within any theater. 19. It is not contemplated that all exploitation will be transferred to Arlington Hall. Before any exploitation now done in any theater is transferred either to another theater or to Arlington Hall Station, it will be determined exactly where such exploitation can be most effectively conducted . For example, such consideration might show that it would be desirable to forward recovered keys in certain systems to an exploitation center within a theater in order that traffic on specific circuits, known to carry material relating particularly to that theater 's operation, might be read more expeditiously. Only by centralized control, based on knowledge of all pertinent facts , plus the capabilities of respective theaters, can intercept and solution problems be most intelligently integrated or can the most effective results be produced. 20. Centralized control of s ignal int elli gence activities cannot properly be characterized as "absentee control." The reason for placing the control in the War Department is to assure unified control which can be provided only by a War Department agency. Even under the existing organization, actual control of signal intelligence activities in SWPA is exercised from distances of 3,500 miles and the Navy controls Guam signal intelligence from Washington , 8,000 miles away. It is immaterial whether the distance is 100 miles or 10,000 miles, provided the control of all activities is exercised by a single agency having authority to coordinate. 21. The comment from SWPA further states that the proposa l would "serve to disrupt an Allied agency that is now functioning most effectively." The Allied agency referred to is the Central Bureau Brisbane, a joint U.S.-British organization under the control of the Chief Signal Officer, SWPA. Considered alone, thi s organization is operated with a relatively high degree of efficiency. However, from the overall viewpoint its operations are wasteful, since they are uncoordinated with other much more extensive signal intelligence operations of both the U.S. Army and the Briti sh. The Briti sh have proposed that the principal cryptanalytic work of the SWPA agency, which is absorbing large numbers of sk illed linguistic and technical personnel, be transferred to Washington where the British believe the work can be done more effectively. There is no basis in the past or presem relations of

220

SIGNALS INTELLIGENCE

e ithe r the War De partment or SWPA with the Briti sh for the suggestion that the proposed centrali zed contro l will have anything other than a highl y sa lutory effect upon cooperation with the Briti sh in the field of signal intelligence. The proposed creation o f a Briti sh Theater in rear of the American zone of o perations in the Pacific will further compli cate the present unsatisfacto ry si tuation. 22 . Probably the most unfortunate si ng le consequence of our present decentralized signal intelligence organizat ion is the strong tendency, most marked in SWPA, for each theater to stri ve for self-sufficiency and to undertake tasks far beyond its capacity to perform. This has at times been carried to ridi culous lengths, and has resulted in (a) useless duplication of effort and waste of highly technical and spec ia li zed person ne l and machinery avai lab le onl y in limited numbers, and (b) neglect of opportunities on the part of theater signal intelligence organizations to derive intelligence of great and immedi ate tactical value from low level cryptanalysis and local traftic analys is. 23. It is certain that centralization of contro l of signal intelligence activities, rather than reducing the amount of useful signal intellige nce wo rk done within theate rs of operations, will free pe rsonnel and units-now wasted on hi gh leve l work beyond the ir capabilities-for greater concentration upon low level sources which can be best ex ploited loca ll y and the possibilities of which have been gross ly neg lected. Thi s will require the fu ll backing and technical support of a central agency having authority ( I) to coordinate the missions of all signal intelligence units including those engaged in low level work, (2) to ensure prompt interchange of technical data and results, and (3) to consolidate results from all sources for both technical and intelligence uses. V. Procedure for placing cel1lralized cOl1lrol inlO effect 24. Pursuant to a directive from the Deputy Chief of Staff dated December 10, 1944 ... , operational command and control of the signal intelli gence activities of the Signal Security Agency are vested in the A. C. of S. , G-2 , WDGS , who exercises such cont rol through the Chief, Military Intelligence Service. This permits close worki ng relations between the production of signal intelligence and its eva luati on, research and dissemination through MIS Special Security Officers attached to the operational headquarters and all units requiring information from thi s source. 25. The Signal Intelligence Units and pe rsonnel now unde r the command of theate r commanders should be placed under the command of the Chief, Military Intelligence Service who wi ll be in a position to integrate the activities of such units and personne l with acti viti es of the Signal Security Agency and to direct all such activities in the light of intelli gence requirements. The transfer of such units to be command of the Chief, Military Intelligence Service, will require no inte rruption in the ex isting working arrangements. The on ly immediate operational effect of the transfer will be that the present Theater Signal Intell igence officers will report directly 10 the Chief, MIS th orugh th e Signal Security Agency, with which they all now communicate directl y on technical matters.

RESHAPING THE TOOLS

221

26. The Signal Intelligence Units and personnel transferred will, as in the case of the Navy 's Advanced Signal Intelligence Units and the Signal Security Agency 's fixed intercept stations, remain under theater commanders for administration and discipline.

***

Appendixes

CHRONOLOGY 1917 - Formation of Code and Cipher Bureau within Military Inteiligence Section, War Department General Staff. 1918 - Code and Cipher Bureau redesignated as MI-8. 1919 - Herbert O. Yardley secretly employed by State and War Departments to operate clandestine cryptanalytic bureau. 1920 - William F. Friedman hired by Signal Corps to compi le codes. 1921 - Yard ley's success in reading Japanese diplomatic code paves the way to major U.S. diplomatic victory at WaShington Peace Conference. 1929 - Yardley 's bureau discontinued; all Army crypto log ic activity transferred to Signal Corps where William F. Friedman becomes first head of the Signal Intelligence Service (SIS). 1931 - Yardley published The American Black Chamber, revealing secrets of U.S. cryptanalysis. 1934 - Provisional Radio Intelligence Detachment organized at Fort Monmouth. N.J.

1935 - Maj. Haskell Allison becomes first Anny officer to head SIS. 1938 - 2d Signal Service Company activated to serve as consolidated intercept ann for SIS. 1940 - SIS team led by Friedman succeeds in breaking the Japanese PURPLE machine cipher by purely cryptanalytic methods. Decrypts, given the codename MAGIC, are given restricted dissemination to a few top polieymakers. 1941 - Pearl Harbor attack; U.S. enters World War II.

SIGNALS INTELLIGENCE

226

1942 - Special Branch of MIS organized. - SIS moves to Arlington Hall , is given mission of intercepting all diplomatic traffic, and is repeatedly redesignated. - major intercept stations established at Vint Hill Farms Station , Virginia, and Two Rock Ranch Station, California. 1943 - SIS successor organization succeeds in so lving first Japanese military (v ice diplomatic) system and achieves final World War II designation as Signal Security Agency. - BRUSA agreement gives Army access to British ULTRA; agreement is made that Arlington Hall will serve as principle center for exploitation of Japanese military communications, and that SSA will participate in exploiting German problem. - Military Intelligence Service (MIS) sets up Special Security Officer system to disseminate product to the theater commanders. 1944 - Special Security Officer (SSO) system expanded to service field commands down to armies and numbered air forces. - Special Branch is dissolved to eliminate excessive compartmentation within MIS. American forces invade Normandy. - MIS is given operational control of SSA. 1945 - Anny-Navy Communications Intelligence Board created to direct all U.S. COMINT operations. - Army Security Agency created to centralize all Anny COMTNT assets.

GLOSSARY The documents included in this collection necessarily contain a number of abbreviations and acronyms (some now obsolete) as well as foreign terms that may be unfamiliar to the general reader. This is a brief guide. A-2 A-3

AAF Abt. AIC

A.C. of S. AEF AFHQ AGCT AGO ALF AIS

Armeegruppe

AR AS ASA ASCZ ASF ASTP

ATC AUS Bde. BEW B.P. Btry.

Air Force intelligence staff Air Force operations staff Army Air Forces (U.S., World War II) AbTeilllflg (German word for troops) Air commodore (British and British Commonwealth Air Force rank equivalent to brigadier general) Assistant chief of staff American Expeditionary Forces (World War I) Allied Forces Headquarters (World War II Med ite rranean theater) Anny Genera l Classification Test Adjutant General's Office Austra lian Imperial Forces American Intelligence Service (a semi-autonomous branch of the Military Intelligence Service in World War II that dealt with Western Hemi spheric affairs) Army group Army regulation Air Staff Army Security Agency Advanced Sector, Communications Zone Army Service Forces Army Special Training Program (World War II program in which selected draftees were allowed to attend college unde r military auspices) Air Transport Command Army of the United States Brigade Board of Economic Warfare Bletchley Park (World War II British COMINT headquarters) Balle ry

228 CIA CBB CBI CG C-in-C; CINC CINCPAC CLRO PEARL CO COMINT COM-Z CP CWT DAF DDMl , BAS DEXTER OF "E" EM ETO ETOUSA EWT FBI FEB Flivo

FY FYI G- I G- 2 G-3 G-4 GAF GC&CS

Genfldm. Genobst. GHQ GI GSC; G.S.c.

SIGNALS INTELLIGENCE Cryptanal ysis Central Bureau, Brisbane China-Bunna-India (theater) Commanding general Commander in Chief Commander-in-Chief, Pacific World War II code word assigned to product deri ved from analysis of medi um-level codes and ciphers Commanding officer Communicat ions intelligence Communicati ons zone

Command post Central War Time Desert Air Force Deputy Director of Military Intelli gence, British Anny Staff (Washington, D.C.) Briti sh tern1 for lower-grade signals inte lligence Direction finding (radio) Enigma Enlisted men European Theater of Operations European Theater of Operations, U.S. Army Eastern War Time Federal Bureau of Investigation Feld Ersatz Balla/ion (Gennan fi eld replacement battalion) Gennan Air Force liaison officer attached to ground units Fiscal year For your infonnation Military personnel staff section Military intelligence Military operations staff sect ion Military suppl y staff section Gennan Air Force Government Code and Cypher School (the World War II British COMINT organization headquartered at Bletchley Park) Generalfeldmarscha/ (Gennan rank of Field Marshal , equivalent to General of the Anny in U.S. rank) Genera/oberst (Gennan rank of colonel general, equivalent to general in the U.S. Anny) General Headquarters Government issue

General Staff Corps

GLOSSARY

HF HF/DF HO

229

Hi gh frequency (radi o communication) Hi gh frequency/direction-finding Designation of various types of truck-mounted she hers used to house mobil e signals intelligence operations in World War II Intelligence and Analysis (section) IA Intelligence Bulletin IB IBM International Business Mac hines Inf. Infantry ISUM Intelligence Summary JA Japanese Army Joint Inte lligence Committee JlC Kana Japanese sy llabary used for foreign ternlS Kitchen police KP Lllfrj70lle Gernlan air fl eet (equ iva le nt to a U.S. numbered Air Force) M- I34A Electromechanical cipher device invented by the Signal Intelligence Service in the 1930s M- 134C Impro ved vers ion of the M-134A used in World War n, familiarl y known as the SIGABA M- 138 Strip ciphe r device introduced by the Signal Intelligence Service Mechanical cipher device invented by the Swedish cryptologist M-209 Boris Hagelin and used to e nci pher U.S. Army tactical communi cations in World War II MAAF Mediterranean Allied Air Forces MAGIC Code word assigned to decrypts of Japanese diplomatic communications MF Medium frequency (radio transmission) MI -8 World War I cryptanalytic unit of Military Intelligence Division Military Intelligence Division, War Department General Staff MID Wo rld War II Military Intelligence Service MIS Military occupational specialty MOS MS ; MS/CXX Briti sh category of high-grade signals intelligence MTO Mediterranean Theatcr of Operations North African Theater of Operations, U.S. Army NATOUSA Noncommissioned offi cer NCO NYPE New York Port of Embarkation Order of Battle O/B Officers Candidate School; Office of the Chief of Staff OCS Office of the Chief Signal Officer (U .S. ArnlY) OCSigO Ol ivc drab OD O IC Officer in charge Oberkommando del' Wehrmaclll (the World War II German OKW Arnled Forces High Command)

230 ONI OP OP-20-G OPD OSCRP OSS OWl PEARL PURPLE PW

PX RAAF RAM

SIGNALS INTELLIGENCE Office of Naval Intelligence Observation post U.S. Navy cryptanalytic organization Operations Division (the U.S . Army 's central command organization in World War II) Officers Signal Corps Replacement Pool Office of Strategic Services Office of War Infomlation World War]] code word assigned to product derived from analysis of low-level codes and ciphers Code name assigned by U.S. to advanced Japanese diplomatic machine cipher Prisoner of war Post exchange Royal Australian Air Force Rapid Analytical Machinery

Rasterschuessel German transposition cipher device

ROMAJI

Rrr

Schwerpul1kl

SCR SCU SHAEF SID SIGINT SIS SLU SNAFU SLOE S.O. SRH SS SSA SSB SSD SSO SSR SSS SWPA T/4

Use of English leiters to transcribe Japanese words Radio/telephone German military tenn used to denote "vital point" for attack or defense Signal Corps radio Special Communications Unit (British units used as communications channels for ULTRA) Supreme Headquarters, Allied Expeditionary Forces Signal Intelligence Division World War II British term for signals intelligence Signal Intelligence Service Special Liaison Unit (British equivalent of Special Security Officer) "Situation nomu,l-all fouled up" Special List of Equipment Special order Special Research History Signal Service Signal Security Agency Signal Security Branch Signal Security Division Special Security Officer Special Security Representative Signal Security Service Southwest Pacific Area (MacArthur'S World War 11 command) Technical sergeant, fourth class

GLOSSARY

TIS TAC TAF T/BA

TIE THQ THUMB TIO TOE ULTRA USAAF USAFFE USATS VHF V-J WAC WIC WD WDGS WEC Wehrkreis WIO

WIT WT/ I

Hy "

231

Technica l sergeant, fifth class Tactica l Air Command Tactical Air Force Table of basic allowances Table of equipment Headquarters hut at Vint Hill Farms School Code word assigned to product derived from radio direction finding Table of organization Table of organization and equipment World War 11 code word ass igned to high-level signals intelligence product (originally ULTRA DEXTER) U.S. Army Air Forces U.S. Army Forces, Far East U.S. Army Transport Service Very high frequency (radio transmission) Victory over Japan Women 's Army Corps Wing commander (Royal Air Force - British Commonwealth rank, equivalent to a colonel) War Department War Department Gene ral Staff Wireless Experimental Centre (India-based British COMINT element) Territorial division of World War 11 German Home Army Warrant officer Wireless telegraphy (radio) Wireless telegraphy intelligence (early ternl for traffic intelligence) British te rm for low-level signals intelligence

A DICTIONARY OF PEOPLE, PLACES, AND TERMS This is a list-by no means exhaustive- of some of the key people, places, and military and cryptologic terms that the reader may encounter in the documents. Arlington Hall - a converted girls' school in Arlington, Virginia, which was taken over by the U.S. government in World War II to serve as the headquarters for the Signal Intelligence Service and its various successor organizations. base line - extent of area over which radio direction sets are dispersed. The longer the base line, the more accurate the bearings. Bissell , Clayton W. - Major General Bissell , an Army Air Forces officer, served as Army Assistant Chief of Staff for Intelligence from 1944 to 1945. Bletchley Park - British estate that served as the wartime home of the Government Code & Cypher School - the Briti sh cryptologic organization. bookbreaking - recovering the underlying "values"- i.e., meanings-of the various code groups within a single code. " buck slip" - referral fon]]. Camp Ritchie - This National Guard installation located in the hill s of Maryland was the site of the An]]y's Military Intelligence Training Center in World War II. cipher - method of secret writing involving the transposition or substitution of individual letters or sy llables in a message. Clark, Mark - General Mark Clark served as commander of the Fifth Army in the Mediterranean Theater of Operations. Clarke. Carter W. - Colonel (later Brigadier General) Carter Clarke served as Chief, Special Branch , Military Intelligence Service from 1942 to 1944. He later

234

SIGNALS INTELLIGENCE

became Deputy Chief of the Military Intelligence Service. and we nt on to command the Arm y Sec urit y Agency. C lass IV Installation - in World War II , a base or camp contro lled by a technica l service, as opposed to Army Ground Forces. code - a method or secret writing in volving the substitution or words and phrases I n a message. code va lues - the meanings assigned to difre re nt wo rds in a give n code. collateral - intelligence derived rrom sources othe r than signals intelligence. colored wa rplans - U.S. AmlY warplans devised prior to World War II. Plans were aimed against individual enemy countries, each of which was assigned a dirrerent color code. The plan to defend against Japanese aggression was Plan Orange. compromi se - a security breakdo w n uncovering a secret to the enemy.

converte r - pre-World War II Army te nn ror a mac hine c iphe r device. cover - deceptive information put out to hide the real source or signals intelligence. c ryptanalysis - the science of breaking codes and ciphers . cryptog rapher - someone who composes codes and c iphe rs. crypto logy - thi s te rm . invented by Friedman, e mbraces the wholefield or secret communications. to incl ude cryptanal ysis, cryptography, and intercept. depth - a cryp tologic teml referring to the existence of a surfi cient numbe r of messages in the same code or cipher key. Fabya n, George - Eccentric phil anthropi st who founded Riverbank Laboratories and was responsible fo r the initiat ion or cryptolog ic stud ies in the United States. Fort Monmouth - A majo r Army S ignal Corps installation in New Jersey. Friedman. William - Originally train ed a s a ge ne ti c ist. Fr ied m a n became America's foremost c rypto logist after George Fabyan hired him to so lve the alleged hidden cipher in Shakespea re's plays that wou ld prove the works to have been wri tten by S ir Francis Bacon. He served as code comp il ati on clerk ror the

PEOPLE, PLACES, AN D TERMS

235

Signal Corps from 1920 to 1930, and was the first head of the Signal Intelligence Se rvi ce. Hi s majo r accompli shm e nt was in leadin g the team that bro ke th e Japanese mac hine c ipher that the United States assigned the codename PUR PLE. During World War II , he served as Director of Commun ications Research for the Signal Securit y Age ncy. "gig" - U.S. Ann y tenn fo r a demerit in an inspection. goniometric - having to do with radi o direction-finding. grades - enli sted ranks. Before World War II , these we nt from Grade 7 (buck pri vate) to Grade I (mas ter sergeant). Halmahera - Island in the Moluccas. indoctrination - in inte lli gence, the procedure whereby selected individual s are instructed on the importance of the secret they will ha ve access to and the need to avoid taking any ac tions that might result in compromise. intercept - in signals intelligence, the act of li stening to broadcast communications. key; keying elements - elements applied to plaintext to produce a ci phered message. King, Ernest J. - Admira l (later Fleet Admiral ) King served as Commander in Chief, United States Fleet and Chief of Naval Operations thro ugho ut World War II. Marshall , George Cat len - General (later General of the Ann y) Marshall served as Anny Chie f of Staff throughout World War I I. McCloy, John J. - a prominent lawyer who served as Ass istant Secretary of War during World War II. McConnack , Alfred - Colonel Alfred McConnack, a Wall Stree t lawyer in ci vilian life, was called into government service to analyze intelligence problems at the begi nning of World War II. He served as deputy chief of the Spec ial Branch until 1944. and the n became Direc tor o f Int e lli ge nce in a reorga ni zed Mil ita ry Intelli gence Service. A fter World War II , McConnack served brie fl y as the first head of the State Department 's Bureau of Intelligence and Research. McNamey, James J. - Lieutenant General Mc Narney served as Ann y Deputy Chief of Staff in World War II and was deeply invo lved in organi zati onal issues.

236

SIGNALS INTELLIGENCE

Miles, Sherman - Major General Shennan Miles was the Anny 's Assistant Chief of Staff for Intelligence at the time of the disastrous surprise at Pearl Harbor. Montgomery, Bernard - Field Marshal Montgomery commanded the British Eighth Anny against Rommel in World War II and later went on to command the 21 Anny Group under Eisenhower. non-indicator traffic - messages sent without any obvious indication of what particular code or cipher key they have been enciphered in. null - radio direction-finding ternl. A "null" occurs when a radio station can no longer be heard, indicating that the antenna of the receiving set is directly aligned on a bearing pointing lowards the emission source.

overhead - personnel not assigned to an organized Anny unit. period - time-frame in which a given code or cryptosystem was allowed to remain in use. Codes and ciphers were changed regularly to protect security. ratings - lower-ranking soldiers before World War II could acquire additional pay within their basic grade by demonstrating professional proficiency in technical specialties. reserve editions (codes) - new codes would be held in reserve to be issued in case existing codes had become compromised. Riverbank Laboratories - This "think tank " was founded by George Fabyan at Geneva, Illinois, to conduct research in a number of scientific field s, ultimately including cryptology.

Scharnhorst - World War II Gennan battle crui ser sunk by Briti sh Navy as a result of successful exploitation of COMINT. system indicators - infonnation at the beginning of a message telling the receiving operator what code or cipher the message has been transcribed in. Stimson, Henry L. - prominent statesman who served as Hoover 's Secretary of State, where he abolished codebreaking aClivity, and later as President Franklin Roosevelt's Secretary of War, where he sponsored it. Strong, George Y. - Major General Strong served as the Anny's Assistant Chief of Staff for Intelligence from 1942 10 1944.

PEOPLE, PLACES, AND TERMS

237

superenciphennent - enciphering a message already rendered into code. Taylor, Telford - Colonel (later Brigadier General) Telford Taylor, a lawyer in civilian life, was commissioned as an intelligence officer in World War II and became the Military Intelligence Service 's chief liaison officer in Britain. After World War II, he became a military prosecutor in the war crimes trials at Nuremberg, Gennany. transposition - a method of enciphennenr in which the order of letters is systematically altered. Two Rocks - Two Rock Ranch near Petaluma, California, was the site of the Signal Security Agency 's main West Coast monitoring station in World War II. Vint Hill Farms - This large estate near Warrenton, Virginia, became the World War II site of the Army's cryptologic training school and Monitoring Station Number I. Yardley, Herben O. - Major Yardley headed up MI-8, the Anny's cryptologic unit in World War l. He later headed up the so-called "Black Chamber" element that broke codes for the State Department and War Department in the 1920s. After being dismissed by Henry L. Stimson, he went on to write a sensational and damaging expose of his activities.